Degradation of Amines - COnnecting REpositories · 2016-04-20 · Degradation of Amines. Maren...

Degradation of Amines

Maren Teresa Johansen

Chemical Engineering and Biotechnology

Supervisor: Anne Fiksdahl, IKJCo-supervisor: Hallvard Svendsen, IKP

Hanna Knuutila, IKP

Department of Chemistry

Submission date: June 2013

Norwegian University of Science and Technology

i

Acknowledgements

This work has been performed at the Department of Chemistry at the Norwegian

University of Science and Technology (NTNU). Some of the analyses have also been done in collabo-

ration with the SINTEF Biotechnology group.

I would like to thank my supervisor, Anne Fiksdahl, for giving me the opportunity to do this thesis,

and Hallvard F. Svendsen for guidance and advice at the Department of Chemical Engineering. I

would also like to thank my co-supervisor Hanna Knuutila for following the process, and offering help

in all practical aspects of the work. Solrun Vevelstad has been very helpful by giving guidance in the

experimental work and for sharing knowledge through scientific discussions.

The support I’ve had from my friends and family has been highly motivational, and given me strength

to both have a social life and a life in the lab.

ii

Abstract

In view of the rising amounts of greenhouse gases in the atmosphere, preventing CO2 emissions has

become increasingly important. The combustion of fossil fuels for energy production and transporta-

tion is a large contributor to the problem.

One of the ways to reduce the amounts of CO2 being released from combustion is carbon capture

and storage (CCS). Post-combustion is the capturing method which has been deemed the easiest to

apply to existing power plants in a short period of time. Absorption of CO2 by MEA is the most com-

mon method used in post-combustion carbon capture, but there are still many aspects of the process

that are not fully understood. Understanding the absorption mechanisms will make it easier to make

more economical and environmentally friendly choices in the future.

In this thesis the oxidative degradation of monoethanolamine (MEA) has been studied using an open

batch setup. The stability of MEA has been studied under different temperatures and concentrations

of oxygen in the gas stream. These experiments give a matrix of experiments performed at 55, 65

and 75 °C, with oxygen concentrations of 6, 21, 50 and 98% in the gas stream. To monitor how well

the experimental results could be trusted, the water balance was maintained throughout the exper-

iments, and the pH was measured in the flasks capturing volatile degradation compounds.

To get a detailed picture of the degradation, the weight percent of nitrogen and the CO2 concentra-

tion has been found in the end samples, and the alkalinity and MEA concentration was found for all

the samples.

11 known degradation compounds have been monitored for the different experiments, and the con-

ditions these compounds are formed at have been compared with the suggested reaction mecha-

nisms. 4 of the products were analyzed as anions using Ion chromatography (IC), and 7 secondary

reaction products were analyzed as part of a degradation mix in LC-MS.

The dependency of these compounds to temperature and oxygen conditions has been discussed.

The primary degradation compounds seems to show a more direct correlation to oxygen flow or

temperature, while the secondary degradation reaction shows a bigger variation of temperature and

oxygen dependency relative to the conditions of the experiments.

Various analytical methods for determination of the known compounds were used to determine the

concentration of the degradation compounds in the experiments. The accuracy of these methods

was investigated, and the results investigated for both LC-MS, GC-MS and IC-EC, showed large varia-

tions.

Mixing experiments were performed to investigate the unknown mechanism of N-(2-hydroxyethyl)

glycine.

iii

Sammendrag

I lys av den økende mengden klimagasser i atmosfæren, har fangst av CO2 blitt stadig viktigere.

Forbrenning av fossilt brennstoff for å mette klodens økende energibehov bidrar til at store mengder

CO2 slippes ut i atmosfæren.

En av måtene for å redusere mengden CO2 som blir frigjort fra forbrenning er karbonfangst og lagring

(CCS). ”Post-combustion” fangst av CO2 er det alternativet som er enklest å ta i bruk for eksisterende

kraftverk. Absorpsjon av monoetanolamin (MEA) er den vanligste metoden som brukes i post-

combustion fangst, men det er fortsatt mange aspekter av prosessen som ikke er forstått fullt ut. Å

øke forståelsen av absorpsjonsmekanismene vil gjøre det lettere å lage mer økonomiske og

miljøvennlige valg i fremtiden.

I denne avhandlingen har den oksidative nedbrytning av monoetanolamin (MEA) blitt studert under

et åpent batch system. Stabiliteten av MEA er undersøkt under forskjellige temperaturer og

konsentrasjoner av oksygen i gass-strømmen. Disse eksperimentene gir en matrise av eksperimenter

utført ved 55, 65 og 75 ° C, med oksygenkonsentrasjoner på 6, 21, 50 og 98% i gass-strømmen. For å

overvåke gyldigheten av de eksperimentelle resultatene ble vannbalansen holdt, og pH ble målt i

gassbolbleflaskene brukt til å fange flyktige degraderingsprodukter.

For å få et detaljert bilde av degraderingen, har vektprosenten av nitrogen og CO2-konsentrasjonen

blitt funnet i sluttprøvene, og alkalitet og MEA-konsentrasjonen ble funnet for alle prøvene.

11 kjente degraderingsprodukter ble vurdert i de ulike forsøkene, og betingelsene for dannelse av

disse produktene har blitt sammenlignet med de foreslåtte reaksjonsmekanismene. 4 av produktene

ble analysert som anioner ved hjelp av ionekromatografi (IC), og 7 sekundære reaksjonsprodukter ble

analysert som en del av en degraderingsproduktblanding i LC-MS.

Avhengigheten av disse forbindelsene til temperatur og oksygenkonsentrasjon i gassstrømmen har

blitt diskutert. De primære degraderingsproduktene synes å vise en mer direkte sammenheng til

oksygenstrøm og/ eller temperatur, mens de sekundære degraderingsreaksjonene viser en større

variasjon av temperatur- og oksygenavhengighet i forhold til betingelsene for forsøkene.

Forskjellige analysemetoder for bestemmelse av forbindelsene ble brukt til å bestemme

konsentrasjonene av degraderingsproduktene i eksperimentene, og det ble funnet usikkerheter i

både LC-MS, GC-MS og IC-EC.

Blandingseksperimenter ble utført for å undersøke den ukjente mekanisme av N-(2-hydroksyetyl)

glycin.

iv

Symbols and Abbreviations

M Molar mass

NL Mass of gas equal to the mass of 1 liter

Vmax Maximum variation in the difference between two results

CCS Carbon capture and Storage

CAS Chemical Abstracts Service

CI Chemical ionization

EI Electron impact ionization

FID Flame ionization detector

GC Gas chromatography

GC-MS Gas chromatography-mass spectrometry

GHG Greenhouse gases

GLC Gas-liquid chromatography

GSC Gas-solid chromatography

HSS Heat-stable salts

IC Ion chromatography

IC-EC Ion chromatography-electrochemical detection

LC Liquid chromatography

LC-MS Liquid chromatography-mass spectrometry

M(n+1)+ Oxidized metal cation

m/z Mass to charge ratio

MS Mass spectrometry

MFC Mass Flow Controller

NIST National Institute of Standard

NPD Nitrogen-phosphorous detector

PLOT Porous layer open tubular

PPM Parts per million (mg/kg)

SIM Selective ion mode

TCD Thermal conductivity detector

WCOT Wall coated open tubular

wt% Weight percentage

α Loading of solution (mol CO2 per mol amine)

α Carbon in α-position

β Carbon in β-position

Ω Electrical resistence

BHEOX N, N’-bis(2-hydroxyethyl) oxalamide

ClO2 chlorine dioxide

CO2 carbon dioxide

DEA diethanolamine

Fe3+ ferric ion

v

H2 hydrogen

H2O water

H2SO4 sulphuric acid

HEA N-(2-hydroxyethyl) acetamine

HEF N-(2-hydroxyethyl) formamine

HEI N-(2-hydroxyethyl) imidazole

HEGly N-(2-hydroxyethyl) glycine

HEPO 4-(2-hydroxyethyl) piperazin-2-one

HEEDA N-(2-hydroxyethyl) ethylenediamine

Na2SO4 sodium sulphate

MEA monoethanolamine

NH3 ammonia

O2 oxygen

O2% Volume fraction of oxygen in gas flow into reactor

OZD 2-oxazolidinone

vi

Table of Contents

ACKNOWLEDGEMENTS I

ABSTRACT II

SAMMENDRAG III

SYMBOLS AND ABBREVIATIONS IV

TABLE OF CONTENTS VI

1. INTRODUCTION 1

2. LITERATURE REVIEW 3

2.1. DEGRADATION OF MEA 3

2.1.1. OXIDATIVE DEGRADATION MECHANISMS 3

2.1.2. DEGRADATION PRODUCTS 7

2.2. ANALYSIS OF DEGRADATION 11

2.2.1. GC-MS 11

3. EXPERIMENTAL 15

3.1. THE OXIDATIVE DEGRADATION SETUP 15

3.1.1. THE OXIDATIVE DEGRADATION APPARATUS 15

3.1.2. CALIBRATION OF THE MASS FLOW CONTROLLERS 17

3.2. THE OXIDATIVE DEGRADATION EXPERIMENT 18

3.2.1. PREPARATION OF MEA SOLUTION 18

3.2.2. EXPERIMENT STARTUP 18

3.2.3. SAMPLING 18

3.2.4. FINISHING THE EXPERIMENT 19

3.3. MIXING EXPERIMENTS 20

3.3.1. LOW PH EXPERIMENT 20

3.3.2. HIGH PH EXPERIMENT 20

3.4. ANALYTICAL METHODS 20

3.4.1. GC-MS 20

3.4.2. IC-EC 21

3.4.3. LC-MS 21

3.4.4. AMINE TITRATION 21

3.4.5. CO2 TITRATION 21

3.4.6. DENSITY, HEAT-STABLE SALTS AND KJELDAHL 22

vii

4. EVALUATION OF UNCERTAINTY IN ANALYTICAL RESULTS 23

4.1. IC-EC 23

4.1.1. INSTRUMENTATION AND METHODS 23

4.1.2. COMPARISON OF SULPHATE CONCENTRATIONS 24

4.2. LC-MS 25

4.3. GC-MS 26

5. RESULTS AND DISCUSSION 29

5.1. VALIDATION OF EXPERIMENTS 29

5.1.1. WATER BALANCE 29

5.1.2. PH IN WASHING BOTTLES 30

5.2. OVERALL DEGRADATION IN EXPERIMENTS 31

5.2.1. NITROGEN BALANCE 31

5.2.2. KINETICS 33

5.2.3. AMINE LOSS 34

5.2.4. LOADING OF CO2 IN END SAMPLES 36

5.2.5. ANION BALANCE 36

5.3. DEGRADATION COMPOUNDS 39

5.3.1. FIRST ORDER DEGRADATION COMPOUNDS 39

5.3.2. SECONDARY DEGRADATION COMPOUNDS (SDC) 43

5.4. MIXING EXPERIMENTS FOR HEGLY 51

6. CONCLUSION 53

7. SUGGESTIONS FOR FURTHER WORK 55

REFERENCES 56

APPENDIX A: EXTERNAL STANDARDS USED FOR QUANTIFICATION OF DEGRADATION PRODUCTS BY

GS-MS AND IC-EC I

APPENDIX B1: CALIBRATION OF MASS FLOW CONTROLLERS II

APPENDIX B2: MASS FLOW CONTROLLER PARAMETERS FOR EXPERIMENTS X

APPENDIX C : MEASURED PARAMETERS FOR GAS-WASHING BOTTLES XII

APPENDIX D: RESULTS FOR QUANTIFICATION OF COMPOUNDS IN SOLUTION XVIII

D1: EXPERIMENT M1: 75 °C, 98% O2 XVIII

D2: EXPERIMENT M2: 75 °C, 6% O2 XX

viii

D3: EXPERIMENT M3: 65 °C, 50% O2 XXIII

D4: EXPERIMENT M4: 75 °C, 50% O2 XXV

D5: EXPERIMENT M5: 55 °C, 6% O2 XXVII

D6: EXPERIMENT M6: 65 °C, 6% O2 XXIX

APPENDIX E: LC-MS DEGRADATION MIX RESULTS FOR MIXING EXPERIMENTS. XXXI

Introduction

1

1. Introduction

The global temperature is rising, and human emissions of greenhouse gases (GHG) are one of the

main contributers [1]. This can have a large effect on the environment, and as of 2009 over 100 coun-

tries have agreed to try to keep the rise in temperature below 2 °C relative to pre-industrial tempera-

tures [2, 3]. CO2 is the largest contributor to the increase of greenhouse gases, and the concentration

of CO2 has been increasing in the atmosphere, as can be seen by the Keeling curve, Figure 1.1 [4].

Figure 1.1: The concentration of CO2 in the atmosphere, measured at Mauna Lua in Hawaii[5].

One of the methods that can be used to minimize the release of CO2 to the atmosphere is carbon

capture and storage (CCS) [6]. CCS uses a range of technologies to capture CO2 emitted from prepa-

ration or combustion of fossil fuels, and from certain industrial processes [7]. These technologies are

separated in three major capture systems: Oxy-fuel, post-combustion and pre-combustion capture,

Figure 1.2 [8]. In addition chemical looping combustion processes (CLC) can be used to separate CO2

directly from the combustion chamber as a separate stream [9].

Figure 1.2: The three main CO2 capture technologies [10].

http://upload.wikimedia.org/wikipedia/commons/1/15/Mauna_Loa_Carbon_Dioxide_Apr2013.svg

Introduction

2

Oxyfuel combustion uses pure oxygen for combustion giving a flue gas consisting of H2O and CO2,

where CO2 can separated easily. Pre-combustion systems process the fuel into CO2 and H2, where H2

is used as the energy source for combustion. Post-combustion systems uses technologies such as

absorption, adsorption and membranes to separate or remove the CO2 from the flue gas after

combustion [11].

Post-combustion carbon capture is the only technique than can efficiently be employed on exiting

power plants, and absorption is the most effective method for large-scale plants[12]. The process

involves absorption of CO2 by a suitable solvent, and the CO2 is subsequently regenerated by heating

the solution in a desorption column, Figure 1.3. Rao and Rubin showed that an amine based absorp-

tion systems are suitable for capturing CO2 in combustion based power plants[13].

Figure 1.3: Flow sheet for amine-based CO2 capture and removal[14].

The most common absorption solvents are amines. The reactions taking place for the capture of CO2

by primary and secondary amines in the abosorber are carbamate and bicarbonate formation. In the

stripper CO2 is regenerated through carbamate reversion. The carbamate formation for the absorp-

tion of CO2 by MEA is given in Equation 1.1.

CO2 + 2NH2CH2CH2OH COO-NHCH2CH2OH + NH2+CH2CH2OH (1.1)

2. Literature review

3


2.1. Degradation of MEA

Three types of amine degradation mechanisms are important in flue gas CO2-capture; thermal, oxida-

tive and CO2 induced degradation[15]. Thermal degradation usually occurs in the absence of O2. The

increased rates of thermal degradation in the presence of CO2 has been explained by aqueous CO2

acting as a proton donor[16]. Oxidative degradation or autoxidation degradation occurs in the pres-

ence of O2. This is the main reaction type occurring in the absorption column in a CO2 capture plant.

CO2 induced degradation is a reaction where CO2 acts irreversibly with amines. This reaction with

primary and secondary amines gives both oxazolidinones and diamines[17, 18].

In this thesis the oxidative degradation of monoethanolamine (MEA) loaded with CO2 has been inves-

tigated.

2.1.1. Oxidative degradation mechanisms

Oxidative degradation in a CO2-capture plant is a result of oxygen in the flue gas reacting with the

amine absorbent. It is believed that the oxidative degradation of amines goes through a radical

mechanism catalyzed by dissolved metals[19, 20]. This radical mechanism is still not fully understood

but there is evidence that it proceeds through an abstraction of hydrogen from nitrogen, α- or β-

carbon, or an abstraction of an electron from the nitrogen lone pair, or a combination of both. Which

mechanism takes place depends on the structure of the amine, the solvents, the concentration, the

acidity and the oxidants[16, 21].

Figure 0.1 describes the different amine autoxidation routes[15, 22]. Route (A) is the direct pathway

where a oxidized metal cation [M(n+1)+] abstracts one electron from nitrogen R, in RH, as the initial

step. This can occurs with both metal and fly ash as a catalyst for the reaction. The metal ion can be

reoxidized by hydroxyl radical formation from oxygen, O2. In route (B) hydrogen is abstracted from

the amine, RH, by a hydroxyl radical or an organoperoxy radical. Reaction (C) occurs by hydrogen

being abstracted from the organoperoxy radical by an amine RH.


4

Figure 0.1: Amine autoxidation mechanism as proposed by Bedell[15].

An example of the electron abstraction mechanism (A) for MEA is given in Figure 0.2[19]. Reaction (I)

is catalyzed by Fe3+, ClO2, R·, or other one-electron oxidants to give an aminium ion[23-25]. The pro-

ton abstraction in (II) can cause reorganization of the molecule giving a carbon-centered radical. Per-

oxide radical is formed when oxygen reacts with an imine radical (III). This leads to the production of

imine and hydrogen peroxide (IV). The imine compound formed by a second reaction with an oxidant

(V) hydrolyses to give ammonia and hydroxyacetaldehyde (VI), whereas formaldehyde and ammonia

are formed during oxidative fragmentation of the imine (VII)[21].


5

Figure 0.2: Degradation of MEA by single electron transfer by mechanism (A), Figure 1.

The alternative hydrogen abstraction mechanism is initiated by an oxidant such as ClO2 [26]. The

mechanism of MEA is proposed by Petryaev et al [27]. The mechanism proceeds through a five-

membered ring transition state, formed by hydrogen bonds between the amine- and alcohol func-

tionality in the alkanolamine. A hydrogen radical is abstracted from the molecule, cleaving the car-

bon-nitrogen bond by a radical chain. Studies conducted by Alejandre et al., Button et al, and

Vorobyov et al. have supported the validity of the cyclic transition state [28-30]. Based on the study

by Petryaev et al. the suggested mechanism is shown in Figure 0.3 [31].


6

Figure 0.3: Hydrogen abstraction mechanism for MEA degradation by mechanism (B), Figure 2.1.

Both the electron- and hydrogen abstraction mechanisms have been supported by kinetic and exper-

imental studies [21]. Hull and Rosenblatt have also found evidence indicating that electron abstrac-

tion is dominant for tertiary amines and hydrogen abstraction is dominant for secondary and primary

amines.

The formation of glycine shown in Figure 0.4 is an example of an MEA radical reacting with oxygen

via a peroxy radical (C). This type of reaction yields both imines and glycines [15].

Figure 0.4: Proposed reaction pathway for production of glycine from MEA, by mechanism (C), Figure 1.


7

2.1.2. Degradation products

Based on the radical mechanisms presented in the previous section, some first order degradation

products have been identified as either fragments or oxidized fragments of the MEA molecule [32].

The aldehydes in Figure 0.5 can be regarded as fragments of MEA, whereas the organic acids are

oxidized fragments of MEA.

Figure 0.5: First order oxidation products.

In addition to the degradation products shown in Figure 5, other known products of oxidative degra-

dation of MEA that can be detected by the current analysis methods are shown in Table 0.1.

Table 0.1: Secondary MEA degradation products.

Name Abb. Structure M

(g/mol) CAS

2-oxazolidinone OZD

1 87.08 497-25-6

N-(2-hydroxyethyl) imidazole HEI

2 112.13 1615-14-1

N-(2-hydroxyethyl) formamine HEF

3 98.09 693-06-1

N-(2-hydroxyethyl) acetamine HEA

4 103.12 142-26-7

4-(2-hydroxyethyl) piperazin-2-one

HEPO

5 144.17 23936-04-1

N, N’-bis(2-hydroxyethyl) oxalamide

BHEOX

6 176.17 1871-89-2

N-(2-hydroxyethyl) glycine HEGly

7 119.12 5835-28-9

The compounds formed as first stage degradation products are reactive chemicals. Da Silva et al.

have proposed that these chemicals can react with solvents or other degradation products. The re-


8

action of MEA or other primary amines with the organic acids formed in Figure 0.5 is believed to lead

to many of the identified secondary oxidative degradation compounds (SDC) [32]. The reaction pro-

ceeds through a nucleophilic attack of the carbonyl group by the primary amine of MEA, as shown in

Figure 0.6.

Figure 0.6: Reaction between MEA and first order degradation organic acids.

The proposed mechanism for the formation of OZD, 1, is shown in Figure 0.7[33]. This reaction is a

type of CO2 induced degradation and is not believed to involve radical intermediates.

Figure 0.7: The mechanism for formation of OZD 1.

The mechanism for the formation of HEI, 2, is not yet confirmed, but the compound has previously

been obtained by a reaction of ammonia, MEA, formaldehyde and glyoxal. The mechanism shown in

Figure 0.8, is proposed by Katsuura et al. and Ben[34-36]. The hetrocyclization mechanism of the two

products is uncertain.


9

Figure 0.8: Hetrocyclization mechanism for the formation of HEI 2.

In the work done by Vevelstad, a new mechanism for the formation of HEI was suggested [37]. This

mechanism is based on the findings of Katsuura, where 2-(methyleneamino)-ethanol and

iminoacetaldehyde was reported to produce HEI [34]. These reactants can be formed from MEA,

formaldehyde, glyoxal and ammonia, which have all been found in MEA degradation solutions [31,

38]. Vevelstad also proved that HEI could be formed from these four compounds by mixing them at

60 °C. The proposed mechanism is given in Figure 2.9.

Figure 2.8: New mechanism proposed by Vevelstad for the formation of HEI, 2, in oxidative degradation.

Da Silva et al. have proposed a reaction mechanism for the formation of HEPO, 5, where HEGly, 7,

reacts with MEA. HEGly is proposed to be formed as a reaction between a first stage oxidative degra-

dation compound and MEA. The precursor of the reaction is however not confirmed [32]. In Section

5.4, a suggested mechanism for the formation of HEGly has been investigated. Figure 0. shows the

proposed mechanism for the formation of HEPO. This reaction mechanism also gives an explanation

for the formation of iso-HEPO. Another proposed reaction mechanism for the formation of HEPO is a

reaction where two MEA molecules react to give N-(2-hydroxyethyl) ethylenediamine (HEEDA).

HEEDA reacts with glycolic acid, and the condensation product cyclizes to give HEPO or iso-HEPO.


10

Figure 0.10: Proposed mechanism for the formation of HEPO, 5.


11

2.2. Analysis of degradation

To understand the composition of the degradation compounds several analytical techniques have

been used[39]. The most common methods for mapping and quantifying degradation compounds

are GC-MS (Gas Chromatography-Mass Spectrometry), LC-MS (Liquid Chromatography-Mass Spec-

trometry ) and IC-EC (Ion Chromatography with Electrochemical detection)[40]. The need for multiple

analysis methods arise from the different functionalities of the degradation compounds including

amines, amides, carboxylic acids, carbonates and carbamates. In this thesis, the use of GC-MS for

identification has been investigated. Titration is a typical method for analysis of the total amount of

amines and CO2 in a sample. The Kjeldahl method is another method used for quantitative determi-

nation of nitrogen in chemical substances.

2.2.1. GC-MS

GC-MS is an analytical method used to investigate multiple compounds in one sample[41]. It is a

combination of two common analytical methods, gas chromatography and mass spectrometry, Fig-

ure 0.9. GC-MS is used for the separation of amines because it was found to be a sensitive technique,

which was able to separate and detect a great number of degradation products within a short period

of time[33].

Figure 0.9: A schematic illustration of the GC-MS instrument.


12

Separation of amines by gas chromatography [33]

In gas chromatography the material under analysis (the analyte) is vaporized before it is transported

through a capillary column by a carrier gas, the mobile phase[42]. The analytes need to be stable at

the temperatures and volatile[43]. The inside of the capillary column is coated with a microscopic

layer of liquid or polymer being the stationary phase. The stationary phase interacts with the com-

pounds in the analyte to different degrees. This makes the compounds elute with different speeds

through the column. The time one compound use to travel through the column is known as the re-

tention time. Polarity and molecular weight affects the retention time of a compound. In a specific

column, the retention time of a specific molecule can be used to identify the molecule when the

conditions are the same.

To get a good separation of amines, the system of parameters shown in Table 2.2, has been pro-

posed [44].

Table 0.2: Important parameters to consider for GC-MS analysis of MEA degradation products.

Considered Type selected Reason for choice

Solvent selection Water The amine solution is already aquous. Other

solvents like methanol may react with the

degradation products.

Column Capillary, medium polar,

water-resistant.

Capillary columns have a higher analysis

speed compared to a packed column.

Injection mode Split-injection Some compounds are in low concentration.

This is preferred over split and on-column

injection because it gives a higher resolution.

Gas carrier Helium Hydrogen gas cannot be used with a mass

spectrometer and nitrogen gas gives a lower

detection

The GC-Columns available is either packed columns or capillary columns. Capillary columns are more

favorable because they have shorter retention times and give narrower peaks. The advantage of a

packed column is that they can tolerate a larger amount of sample. Capillary columns are the most

common columns found today [45]. The wall capillary columns are a fused silica matrix. Fused silica is

preferred because of its inertness.

There are two types of capillary columns. The Wall Coated Open Tubular (WCOT) column is a Gas-

Liquid Chromatography (GLC) column whose walls are coated with a liquid stationary phase. Porous

Layer Open Tubular (PLOT) columns coated with a solid stationary phase are Gas-Solid Chromatog-

raphy (GSC) columns. PLOT columns are not used expect for in special cases, for example for separa-

tion of isomers and fixed gases [46].

The usual detectors for gas chromatography are Thermal Conductivity Detector (TCD), Flame Ioniza-

tion Detector (FID), or Nitrogen-Phosphorous Detector (NPD). These detectors have different sensi-


13

tivity, and information. In GC-MS the mass spectrometer acts as the detector, giving both quantita-

tive and qualitative information.

Identification by mass spectrometry [47, 48]

When compounds have been separated through the gas chromatography column, they can be ana-

lyzed in a mass spectrometer. Figure 0.10 shows a schematic figure of how the mass spectrometer

works. The compounds are first ionized before they are accelerated through a magnetic field that

separate the ions based on their mass/charge ratio (m/z).

Figure 0.10: Schematic illustration of mass spectrometer[49].

The ionization method is chosen on behalf of the compounds which are analyzed. The common

methods for gas-phase ionization are Electron Impact Ionization (EI) and Chemical Ionization (CI).

Electron Impact Ionization (EI) is the most common and usually the standard form of ionization used

for GC-MS. When the molecules have entered the MS source, they are bombarded with free elec-

trons emitted from a filament, see Figure 10. When the molecules are attacked by electrons, they

produce radical cations, known as molecular ions. A molecular ion can fragment in a characteristic

and predictable way. The recorded fragment mass specters can be compared with a library which

contains specters of millions of compounds. One of the most common libraries is the National Insti-

tute of Standard (NIST) library. EI Mode is an aggressive ionization method that causes the molecules

to fragment extensively, making it sometimes difficult to find the original molecular mass (the mo-

lecular ion peak).

Chemical Ionization (CI) is the most important type of “soft ionization” techniques used when a mo-

lecular ion peak is desired. Instead of bombarding the molecules with high energy electrons, a rea-

http://upload.wikimedia.org/wikipedia/commons/0/0d/Mass_Spectrometer_Schematic.svg


14

gent gas (methane, isobutene or ammonia) is ionized. The sample molecules collide with the reagent

gas ions at a relatively high gas pressure, and are ionized by proton transfer [M+1]+, electrophilic

addition, [M+18]+, or charge exchange [M]+.

The mass spectrometer can be used in two different modes, Full Scan and Selective Ion Mode (SIM).

These modes can be run separately or combined.

The Full Scan mode gives a complete Mass Spectrometry specter showing all molecular ion peaks.

This is useful for determining unknown compounds in a sample. The Full Scan mode is often used to

identify the specific mass fragments for a molecule before using the SIM method. The target range

usually shown in Full Scan is typically between 50 m/z to 400 m/z.

In Selective Ion Mode (SIM), specific mass fragments of compounds, usually determined from full

scan, are preselected before analysis. When the sequence is run only the chosen mass fragments are

detected in the machine. This makes it easier to detect small fragments since the Mass Spectrometer

only is focusing on a small number of fragments. In SIM mode it is possible to identify all the com-

pound peaks or only focus on one. Quantification of peaks is usually done in SIM mode.

Experimental

15

3. Experimental

3.1. The oxidative degradation setup

3.1.1. The oxidative degradation apparatus

For this thesis six oxidative degradation experiments were run in three parallels. These experiments

will be named M1 to M6 according to when they were run. Two similar apparatuses were used for

the degradation experiments. An illustration of the oxidative degradation experimental apparatus is

shown in Figure 3.1. Oxygen gas, or an air and oxygen gas mix 1, and CO2 gas 2, entered the appa-

ratus through mass flow controllers (MFC) 3. The flow of the different gases was set by the MFC’s,

giving a specific volume percentage of oxygen in the gas stream. This will later be referred to as oxy-

gen percentage or oxygen concentration. The gas streams were later combined to give the reaction

gas stream.

Figure 3.1: Oxidative degradation apparatus.

The reaction gas was bubbled through a reaction vessel 4, which was filled with water to pre-saturate

the gas. This was to maintain the water balance in the main reactor 6.The reaction vessel was cooled

with tap water. The pre-saturated gas and recirculation gas was led into a gas pump (ATB, 50L/min)

5, which pumps the gas into the main reaction vessel 6. The reaction gas was bubbled through the

Experimental

16

prepared MEA solution during stirring. The main reaction vessel was heated with water from a heat-

ing circulator at a set temperature, 55 or 65°C, entering through 14 and exiting through 15. The gas

was then led though two connected condensers cooled with tap water entering through 12 and exit-

ing through 13. The condensers were connected to a two-necked separatory funnel 9. After the sepa-

ratory funnel the gas was led through two gas-washing bottles 10 and 11, filled with H2SO4 (1 M). The

washed gas was led to a vented fume hood.

The set up of the mass flow controllers are given in Figure 3.2. The two apparatuses were run in

three parallels, giving a total of six experiments. Setup 1 was used for the first two experiments,

while the setup 2 was used the last four experiments. For a more detailed description of each exper-

iment see section 3.3.2. MFC 1 and MFC 3 adjusted the mass flow of CO2, and MFC 2 and MFC 4 ad-

justed the mass flow of the premixed oxygen/nitrogen gas into the reactors. The mass flow through

MFC 5 was combined with the mass flow through MFC 6 or MFC 7 and the flows were adjusted to

give the right oxygen percentage for the experiments. In setup 2 the gas mixture controlled by MFC 5

and MFC 7 was used for both reactors.

Figure 3.2: The different mass flow controller setups used for the conducted experiments.

Experimental

17

3.1.2. Calibration of the mass flow controllers

MFC 6 and 7 were calibrated before experiment M1 and M2. All the mass flow controllers except

MFC 7 were calibrated before experiment M3 and M4 was started. This was done by measuring the

amount of nitrogen gas let though the controllers at various opening percentages. Soap solution was

put in a rubber container at the bottom of a glass cylinder of known volume. When the gas was led

into the bottom side of the cylinder, the rubber container was squeezed so that the soap solution

covered the gas inlet for a short time, and a soap film started moving up the cylinder. The time was

taken for the film to move a specific distance corresponding to a known volume.

The openings required for the wanted oxygen mass flow concentrations were calculated using the

obtained calibration curves. The values used for the calibration of each of the mass flow controllers

are given in Appendix B. The openings of the mass flow controllers are given in Table 3.1. The oxygen

and carbon dioxide percentages and volumetric gas velocities for the experiments are given in Table

3.2. For the first two experiments the oxygen percentages were recalculated according to the newest

calibration.

Table 3.1: Percentage openings for mass flow controllers for each parallel of experiments.

Experiment Opening (%)

MFC 1 MFC 2 MFC 3 MFC 4 MFC 5 MFC 7 (6)

M1 and M2 3.4 6.8 3.4 6.3 13.3 8

M3 and M4 2.8 6.1 3.5 6.6 16 12.6

M5 and M6 2.8 6.1 3.5 6.6 29.9 1.7

*MFC 6 was used for experiment M2.

Table 3.2: Flow of CO2 through MFC 1 and 3, and O2 mix through MFC 2 and 4. Total gas flow and total O2 and CO2 vol-

ume percentage for all experiments.

Experiment Flow O2 Mix (NL/min) Flow CO2 (NL/min) Total gas flow (NL/min) O2 (%) CO2 (%)

M1 0.39 0.009 0.4 98 2.3

M2 0.33 0.0073 0.34 5.6 2.1

M3 0.35 0.0075 0.36 49 2.1

M4 0.35 0.0075 0.36 49 2.1

M5 0.35 0.0075 0.36 5.9 2.1

M6 0.35 0.0075 0.36 5.9 2.1

Experimental

18

3.2. The oxidative degradation experiment

3.2.1. Preparation of MEA solution

Water (1400 g, 77.71 mol) was added quickly to a solution of MEA (600 g, 9.82 mol) and stirred at

room temperature for 15 minutes to give MEA (30wt%). CO2 (0.4 equiv.) was loaded to the 30wt%

MEA solution (1 equiv.) by bubbling the gas into a given amount of MEA solution. Na2SO4 (2-1 g) was

added to some of the batches. The specific amounts for each batch are given in Table 3.3.

Table 3.3: Amounts of reagents in each batch.

Experiment MEA 30wt% [g] MEA [moles] CO2 [g] CO2 [moles] Loading, α Na2SO4 [g]

M1 961.09 4.72 83.12 1.889 0.400 1.65

M2 959.85 4.714 83.03 1.887 0.400 1.83

M3 946.23 4.647 81.71 1.857 0.399 -

M4 945.97 4.646 81.68 1.856 0.399 -

M5 952.25 4.677 81.41 1.850 0.396 0.99

M6 961.10 4.721 82.78 1.881 0.398 1.02

3.2.2. Experiment startup

The MEA solution (30wt%, α=0.4) was added to the main reaction vessel. H2SO4 [0.5 M] was added to

the gas-washing bottles. The relevant temperature was set for the heating circulator. The cooling

water was turned on slowly. The gas bottles were opened and the mass flow controllers were turned

on to obtain the chosen conditions. The circulation pump and the magnetic stirring were turned on.

The specific parameters for each experiment are given in Table 3.4.

Table 3.4: Parameters for start-up of the oxidative degradation experiment.

Experiment M1 M2 M3 M4 M5 M6

MEA solution [g] 1044.21 1042.88 1027.94 1027.65 1033.66 1043.88

Temperature [oC] 75 75 65 75 55 65

Oxygen [%] 98 6 50 50 6 6

3.2.3. Sampling

Samples were taken from the reactor and the gas-washing bottles. Sampling time and the sample

mass for the main reaction vessel are given in

Experimental

19

Table 3.5.

Table 3.5: Amount of solution taken from the reaction vessel for each sample.

Exp. M1 M2 M3 M4 M5 M6

Sample

No.

Time

[days]

Sample

mass

[g]

Time

[days]

Sample

mass

[g]

Time

[days]

Sample

mass

[g]

Time

[days]

Sample

mass

[g]

Time

[days]

Sample

mass

[g]

Time

[days]

Sample

mass

[g]

1 0 54.44 0 53.7 0 27.11 0 26.96 0 29.51 0 29.37

2 0.9 4.82 0.9 4.7 1.1 4.9 1.1 4.87 1 4.94 1 5.19

3 3 4.89 3 4.81 3.2 4.97 3.2 4.93 3 4.93 3 5.08

4 6.3 4.89 6.4 4.85 7.1 4.85 7.1 5.08 7 5.03 7 5.03

5 8.1 4.91 8.1 4.81 10 5.09 10 5.21 13.9 4.93 13.9 4.98

6 10.1 4.81 10.1 4.84 14 4.79 14 4.82 20.9 5.29 20.9 5.34

7 13 5.04 13 5.07 21.2 5.07 21.2 4.86 28 5.02 28 4.94

8 17.2 4.71 17.2 4.95 28.1 56.85 28.1 54.95 34.8 4.86 34.8 5.03

9 21.2 4.57 21.2 52.13 41.8 53.85 41.8 55.82

10 24 4.77

11 29.9 4.78

12 35.9 58.78

The gas-washing bottles were refilled with H2SO4 [0.5 M] when samples were taken from the reaction

vessel. The first bottle, 10, was filled with about twice as much H2SO4 as the second bottle, 11. The

sampling time and the mass of H2SO4 added to each bottle in the six experiments are found in Ap-

pendix C.

3.2.4. Finishing the experiment

After 3-6 weeks the experiment was completed. The last samples were taken and the pump and

heating were turned off. The gas bottles for nitrogen, CO2 and O2 were closed. The gas-washing bot-

tles were removed and replaced with a gas-washing bottle with distilled water. The solution in the

main reaction vessel was cooled to room temperature and the solution was removed from the reac-

tion vessel. The reaction vessel was filled with distilled water, stirred for an hour and emptied twice.

The reaction vessel was filled with distilled water, heated to the reaction temperature and stirred

overnight. The reaction vessel was cooled to room temperature, emptied and filled with distilled

water. The cooling water was turned off.

Experimental

20

3.3. Mixing experiments

Two different tests were performed to investigate the formation mechanism of N-(2-hydroxyethyl)-

glycine (HEGly) by preparing mixtures of intermediate compounds assumed to participate in the sug-

gested mechanism.

3.3.1. Low pH experiment

MEA (1.51 g, 0.025 mol) was added slowly to a mixture of formic acid (23.59 g, 0.51 mol) and

glyoxylic acid (50 wt% in water, 7.05 g, 0.048 mol). The reaction mixture was stirred for 72 hours at

40-55 °C. Samples were taken during the experiment, and at the end of the experiment.

3.3.2. High pH experiment

MEA (28.55 g, 0.48 mol) was added slowly to a mixture of water (38.02 g, 2.11 mol) and glyoxylic acid

(50 wt% in water, 2.97 g, 0.02 mol). Formic acid (4.60 g, 0.10 mol) was added, and the mixture was

stirred for 44 hours at 45-55 °C. Samples were taken at the beginning, during and at the end of the

experiment.

3.4. Analytical methods

Different analytical methods were used in order to identify and quantify degradation compounds in

the experiments. The chromatographic procedures for GC-MS, LC-MS/MS and IC-EC will be described

first. The amine concentration and the total amount of CO2 were analyzed by the titration methods

described in the middle of this chapter. Total nitrogen concentration (Kjeldahl), Heat-Stable Salts

(HSS) and density was analyzed by an external contractor.

3.4.1. GC-MS

GC-MS was used to compare the results for HEF and HEPO given by LC-MS.

The MEA solution samples were analyzed on a GC-MS in both Full Scan and SIM mode. The samples

and the standards were diluted with distilled water. The samples were diluted to 100 mg/g. The

standards were diluted to the following concentrations: 2000, 1428, 1000, 750, 500, 200, 100, 70, 50

and 10 µg/g. The carrier gas was Helium.

The analytes were injected into the column by an Agilent 7693 Autosampler. The Varian Woot Fused

Silica, CP-Sil 8CB for Amines, P/N: CP7596, was used to separate the different analytes for SIM analy-

sis of the samples. The column was 30 m with a diameter of 0.32 mm, and was run on a 7890A GC-

system. Quantification was done using the standards listed in Appendix A.

The mass spectrometer was an Agilent 5975C inert XL EI/CI MSD with a Triple-Axis Detector. The

method described by Lepaumier[50] was used for quantification in the beginning. The samples were

then diluted in methanol, see section 4.3.

Experimental

21

3.4.2. IC-EC

Gøril Flatberg at IKP, NTNU performed the IC-EC analyses.

The MEA solution samples were analyzed for anions on a Dionex ICS-5000 IC-EC system. The samples

and standards were diluted with deionized water (18.2 MΩ). The unknown samples were diluted

from 1:50 to 1:10 000 based on mass. The external standards were diluted to 10, 5, 2.5, 1 and 0.5

mg/ml.

The eluent was potassium hydroxide diluted in deionized water (18.2 MΩ) purified by an ICW-3000

water purifier. The system includes an Autosampler (AS-50), a Gradient pump (GP50), a Column Oven

(TCC- 3000), a CD conductivity detector, a CRD200 carbonate removal device, an eluent generator

system, a ASR300 suppressor, and an Ion Pac AS11-HC column (2mm*250mm) with an Ion Pac AG11-

HC guard column (2mm*50mm). Quantification was done using the standards listed in Appendix A.

This method was used to find the concentration of formate, oxalate, nitrate, nitrite, and sulphate

(PPM).

3.4.3. LC-MS

Kai Vernstad and Astrid Hyldbakk at SINTEF biotechnology performed the analyses of the samples on

LC-MS/MS.

The MEA solution samples and the samples from the gas-washing bottles were analyzed on a LC-

MS/MS system. The eluent was formic acid in water (25 mM), with a flow rate of 0.6 ml/min. The

analyte was injected to a Supelco Analytical Ascentis® Express RP-Amide HPLC Column, 2.7 μm, Cat.

No.:53931-U, by an Agilent Infinity Autosampler 1200 Series G4226A. The column was 15 cm with an

inner diameter of 4.6 mm. The molecules were converted to ions by electrospray ionization (ESI). An

Agilent 6460 Triple Quadrupole Mass Spectrometer coupled with an Agilent 1290 Infinity LC Chro-

matograph was used to obtain the mass spectrum.

This method was used to determine the MEA concentration (mol/L) and the concentration of the

degradation products (µg/ml) listed in Table X, section X.X. Some of the samples were also analyzed

for: DEA (mmol/L), nitrosodiethanolamine (NDELA), nitroso-HeGly, NH3 and the alkylamines: methyl-

amine, ethylamine, dimethylamine and diethylamine.

For more details around the method used see study by de Silva et al[32].

3.4.4. Amine titration

The MEA solution samples were analyzed for total amine concentration by titration. The sample (0.2

g) were diluted with distilled water ( 50 ml). The solution was titrated with sulfuric acid (0.2 N) on a

Mettler Toledo G20 Compact titrator equipped with a Mettler Toledo Rondolino carousel and a Met-

tler Toledo DGi115-SC pH probe.

3.4.5. CO2 titration

The amount of CO2 in the first and last sample of the MEA solution for each experiment was analyzed

by titration. The sample (1g) was added to a solution of barium chloride (0.1 N, 25 ml) and sodium

hydroxide (0.1 N, 50 ml). An identical solution without sample was made as a blank sample. The mix-

tures were boiled for 4 minutes, and the precipitate was filtrated. The filtrate was added distilled

water (~40 ml) and HCl (0.1 N, 60 g). This solution was titrated with sodium hydroxide (0.1 N) on a

Experimental

22

Metrohm 809 Titrando. The system includes a Metrohm 814 Sample processer, a Metrohm 800

Dosino dosing unit and a Metrohm 6.0262.100 pH probe. This method is also described by

Ma’mun[51].

3.4.6. Density, Heat-Stable Salts and Kjeldahl

The density was analyzed on a Mettler Toledo DM 40 Density Meter, at 22°C. This was done for the

start and end samples.

The amount of heat-stable salts was analyzed by Merete Wiig at SINTEF using an ion-exchange based

wet chemistry method and titration with NaOH.

The Kjeldahl method was used to analyze the samples for total nitrogen. This was done by an exter-

nal contractor by the standard method[52].

Evaluation of uncertainty in analytical results

23

4. Evaluation of uncertainty in analytical results

4.1. IC-EC

4.1.1. Instrumentation and methods

For the Ion Chromatography analyses, the Instrumentation had been out of order until one month

before the delivery date of this thesis. The method used for analysis of the samples from the reaction

vessel was therefore not optimized.

Because of the short time most of the results for the compound concentrations were generated us-

ing an automated computer method which is less accurate than a more time consuming manual

method. The program had multiple ways of integrating the peaks, and did not respond well to nega-

tive peaks, Figure 4.1. The retention time of the formate peak varied between 8.1 -9.9 min and the

peak were sometimes as broad as 3 min. This led the program to have difficulties recognizing the

peak, and splitting it into multiple integration areas.

Each sample was analyzed using three different dilutions. The automatic program showed higher

concentrations of anions for the most diluted samples. The generated results for experiment M1 are

given in Appendix D1. As an example: The dilutions 1:10 000, 1:1 000 and 1:500 for nitrite in sample

9, gives respectively these results for nitrite: 3689.2, 769.9 and 646.6 PPM. The effect the dilution has

on the computer calculated results seems to mainly influence the anions found in low concentra-

tions. The values that were used in Section 5.2, are mostly the automatically generated results from

the least diluted samples, but some values are from higher dilutions. Therefore, the IC-EC results can

only be used to identify very distinct trends in the anion concentrations by comparing the different

experiments.

For the manual method the integrations were done by hand, but it was not possible to address the

problems concerning broad peaks and an uneven baseline. To do this the samples needs to be reana-

lyzed using a new method of chromatography.

At the end of the work for this thesis most of the samples had been analyzed manually, and these are

the results that will be presented in Section 5.3.1.

Figure 4.1: Figures of IC-EC specters for samples at different dilutions. The colored lines show the automatically generated

areas used to quantify each peak.


24

4.1.2. Comparison of sulphate concentrations

In experiment P1-P4, M1, M2, M5 and M6 known amounts of sodium sulphate (Na2SO4) was added.

This was compared with the IC-EC results for the first sample of each experiment. The values are

given in Table 4.1.

Table 4.1: The sulphate concentration of the first samples calculated from the added amount of sodium sulphate, and the

values given by IC-EC analysis.

Experiment Added sulphate (mg/kg) Measured sulphate (mg/kg) Diff (%)

M1 1068.74 1067.02 0

M2* 1186.84 1270.96 7

M5* 647.79 767.08 18

M6* 660.89 775.83 17

P1 1969.30 1820.68 -8

P2 1930.44 1858.77 -4

P3 1926.22 1608.41 -16

P4 2003.39 1716.24 -14 *Results are calculated from computer program described in section 4.1.1.

These results show that the results given by the computer method described in Section 4.2 give a

higher concentration of sulphate than the actual values. The method used to find the concentrations

for the project specialization gives lower concentration.

For experiment M1 the standard deviation was calculated for the analyzed concentration. Table 4.2

shows the results and standard deviation with the maximum and minimum values for the sulphate

concentration. The sodium sulphate was added to monitor the water balance throughout the exper-

iment. Since the values between minimum and maximum overlaps for all the samples except sample

6 and 8, these results cannot be used to monitor the water balance, and a more precise method of

analysis is needed.

Table 4.2: Sulphate concentrations results with standard deviation from IC-EC.

Sample # Time (Days) Mean PPM) STD (PPM) Min (PPM) Max (PPM)

1 0.0 1067.022 #N/A #N/A #N/A

2 0.9 1096.369 #N/A #N/A #N/A

3 3.0 1294.824 382.600 912.223 1677.424

4 6.4 1263.601 158.101 1105.500 1421.702

5 8.1 1211.233 72.387 1138.847 1283.620

6 10.1 1111.646 40.025 1071.621 1151.672

7 13.0 1135.226 46.948 1088.279 1182.174

8 17.2 1187.511 31.854 1155.657 1219.365

9 21.2 1152.115 23.386 1128.730 1175.501


25

4.2. LC-MS

The MEA concentrations (mol/L) for experiment M5 and M6 were analyzed in parallel (run 1; M5 and

M6 versus run 2; M5 and M6) by LC-MS twice. In the first run the dilutions were done by SINTEF

Biolab, and in the second run the first dilution was done by the author. The results from the first and

second run are compared for both experiments in Table 4.3.

Table 4.3: Results for the first and second analysis of MEA concentrations in M5 and M6.

Experiment M5 M6

Time (days) MEA, run 1

(mol/L)

MEA, run 2

(mol/L)

Difference

(%)

MEA, run 1

(mol/L)

MEA, run 2

(mol/L)

Difference

(%)

0.0 6.08 5.35 14 5.87 5.20 11

1.0 6.24 5.12 22 5.36 4.91 8

3.0 5.97 5.18 15 5.31 4.95 7

7.0 5.82 5.09 14 5.14 4.69 9

13.9 5.50 4.64 18 5.41 4.53 16

20.9 5.39 4.55 18 5.10 4.96 3

28.0 5.02 4.35 15 4.83 4.63 4

34.8 5.10 4.25 20 4.87 4.17 14

41.8 4.92 4.12 19 4.42 4.49 -2

As can be seen from table 4.3, the maximum-minimum difference between the values for run 1 vs

run 2, Vmax, given for the MEA concentration is 24%. It seems like it is a systematic deviation between

the first and second run. These results can therefore not be used to say anything specific about the

reduction of MEA in the samples. The reaction rate for MEA in the experiments will be shown in sec-

tion 5.X. For experiment M5 the Vmax is 7%, this means that the reaction rate for this experiment is

more likely to be close to the actual value compared to experiment M6 which has a Vmax of 18%.

In Table 4.4, the concentrations of MEA (mol/kg) prepared in the initial solutions, before and after

loading with CO2, are compared to the results given for the first sample by LC-MS analysis. The values

for the unloaded solution were used because the CO2 in the samples are released at temperatures

above 121 °C [53]. The measuring method for MEA seems to give higher values than expected. If the

values for each particular experiment are subsequently higher for each sample this will not greatly

affect the calculated reaction rates. The nitrogen balance presented in section 5.2.1 might however

show a higher recovery rate (RN) since this is calculated from the measured MEA values.


26

Table 4.4: Difference in percent between calculated and measured concentration of MEA in start samples.

Experiment

Concentration of MEA

in loaded solution

(mol/kg)

Concentration of

MEA in unloaded

solution (mol/kg)

MEA measured

by LC-MS

(mol/kg)

Difference Unload-

ed/Measured MEA

(%)

M1 4.52 4.91 5.06 3

M2 4.52 4.91 5.77 15

M3 4.52 4.91 5.75 15

M4 4.52 4.91 5.93 17

M5, run 1 4.52 4.91 6.08 19

M5, run 2 4.52 4.91 5.35 8

M6, run 1 4.52 4.91 5.87 16

M6, run 2 4.52 4.91 5.2 6

P2 4.53 4.91 5.22 6

P3 4.55 4.91 5.25 6

P4 4.53 4.91 5.22 6

4.3. GC-MS

The method described by Lepaumier [50] was used to separate the analytes for GC-MS. This method

separated the peaks of the standards, but the signal response for the different dilutions did not cor-

respond well, see Figure 4.2. As can be seen from the figure, the signal for 750 μg/ml was a lot higher

than the signal for 1 000 μg/ml.

This method has been used previously by Vevelstad and Lepaumier without troubles. The instrumen-

tation was then expected to have caused the problem. The old column was replaced without any

results. The pumps were later replaced without improving the signals.

(a) (b)

Figure 4.2: Figure (a) shows the peaks of the standards in a 1500 μg/ml dilution. Figure (b) is a diagram of the HEF peaks

for all the dilutions overlayed.


27

The dilutions had been done with distilled water as described in the method, but since water is

known to be difficult to use in GC columns[54], the standards were diluted in methanol. When

the results for the different methanol dilutions of the standards were compared, the peak areas gave

a better match with the dilution concentrations. This is presented in Figure 4.3 with the calibration

curve for HEF.

(a) (b)

Figure 4.3: Figure (a) shows the HEF peaks of all the methanol dilutions overlaid. Figure (b) shows the calibration curve

obtained from the results.

The product samples for experiment M1 were diluted in methanol, with dilutions 1:20 and 1:50.

Sample 9 was analyzed four times for both dilutions to compare the responses. The other samples

were run twice. The overlaid specters for sample 9 in both dilutions are shown in Figure 4.4.

(a) (b)

Figure 4.4: Overlaid specters of four analyses of the same sample. In (a) the peaks for the 1:20 dilution is showed, and the

1:50 dilution in (b).

0

2.00e+006

4.00e+006

0 1.00e+003 2.00e+003

HEF

Response

Concentration

27.60 27.80 28.00 28.20 28.40 28.60 28.80 29.00 29.20 29.40 29.60 29.80 30.00 30.20 30.400

50000

100000

150000

200000

250000

300000

350000

400000

450000

500000

550000

600000

650000

T ime-->

Abundance

TIC: 1_20 MeOH F1P8.D\ data.msTIC: 1_20 MeOH F1P8 2.D\ data.ms (*)T IC: 1_20 MeOH F1P8 3.D\ data.ms (*)T IC: 1_20 MeOH F1P8 4.D\ data.ms (*)

27.50 28.00 28.50 29.00 29.50 30.000

20000

40000

60000

80000

100000

120000

140000

160000

180000

200000

220000

240000

Time-->

Abundance

TIC: MIX 10 ugmL MeOH2.D\data.msTIC: 1_20 MeOH F1P9.D\data.ms (*)

TIC: 1_50 MeOH F1P9 2.D\data.ms (*)TIC: 1_50 MeOH F1P9 3.D\data.ms (*)TIC: 1_50 MeOH F1P9 4.D\data.ms (*)


28

For the 1:50 dilution both the shape of the compound peaks and the similarity of the responses was

improved. The standard deviation was calculated for the results from the 1:50 dilution, and the val-

ues with standard deviation are shown in Appendix D1.

The results were compared with the results obtained from LC-MS, and the obtained concentrations

for both methods of analysis are shown in Figure 4.5, for HEF and HEPO.

(a) (b)

Figure 4.5: Concentrations for HEF (a) and HEPO (b) analyzed by GC-MS and LC-MS for the samples in M1.

The concentrations found for HEF by GC-MS was up to twice of what LC-MS showed, while for HEPO

the values were up to four times as high. The standard deviation for the GC-MS samples was calcu-

lated to reach 50% of the mean value. The values given by LC-MS fall significantly below the confi-

dence interval calculated for GC-MS, and the values for GC-MS and LC-MS is not comparable. Be-

cause of the discussed problems with GC-MS, the LC-MS results will be deemed the most valid. The

degradation pattern shown for HEF in GC-MS is although similar to the results given by LC-MS, and

maybe GC-MS will give useable results for the samples if they were more diluted.

Results and discussion

29

5. Results and discussion

In this chapter the results obtained from the oxidative degradation experiments will be presented

and discussed. The analytical results for the six experiments performed for this thesis is presented in

Appendix D. The results of this thesis will also be compared with the results given in the project spe-

cialization report (project) [55], and the doctoral thesis of Vevelstad [37]. The combined results give a

matrix of different temperatures and oxygen flow percentages, see Table 5.1. The experiments per-

formed for the project and master thesis, are marked P and M respectively. The experiments per-

formed by Vevelstad are marked V.

Table 5.1: Matrix of oxidative degradation experiments performed at NTNU.

Temperature Volume percentage of oxygen in gas flow

6 % 21 % 50 % 98 %

55 C M5 P2, V1, V2 V3 V4

65 C M6 P4 M3 P3

75 C M1 V5 M4 M2

The results describing the validity of the results for each experiment are presented first. Then the

different experimental results will be compared. The trends shown for the degradation compounds

will be discussed compared to the suggested reaction mechanisms. At the end of this chapter the

results from the mixing experiments will be discussed.

5.1. Validation of experiments

5.1.1. Water balance

The water balance was calculated from the mass of solution added to the reactor at the start of the

experiments minus the total mass of the samples, and the total mass of solution taken out of the

reactor at the end of the experiments. The water balance for each of the experiments is given in

Table 5.2.

The experimental results are considered valid when the change in water mass is less than 7% in ei-

ther direction. Experiment P1 has a water loss of 18% and will therefore not be discussed further in

this thesis.

The increase in water in some of the experiments is due to the wetting of the CO2 gas let into the

reactor. The mass of water absorbed by the CO2 was adjusted by the height of water over the sinter

in the wetting chamber. For experiment M5 and M6, the water was 1 cm above the sinter, giving the

best results for the water balance.

The experiments that show a decrease in water mass has either not had enough water over the sin-

ter in the wetting chamber, or had a leakage in the system somewhere.


30

Table 5.2: Water balance for all oxidative degradation experiments.

Experiment T (°C) O₂ (%) Water balance (%)

M5 55 6 1.5

P2 55 21 -3.8

V1 55 21 7.5

V2 55 21 6.1

V3 55 50 7.5

V4 55 98 6.1

M6 65 6 1.0

P1 65 21 -18.2

P4 65 21 -3.8

M3 65 50 -2.2

P3 65 98 -5.5

M1 75 6 7.1

V5 75 21 6.0

M4 75 50 -1.1

M2 75 98 -2.2

5.1.2. pH in washing bottles

The exit gas was bubbled through two flasks containing H2SO4 (0.05 M) to capture ammonia and

MEA. The flasks were changed every 2-7 days, depending on the assumed amount of degradation.

To investigate if the maximum capture capacity was reached for any of the experiments, the pH in

the washing bottles were measured. The results are given in Table 5.3. Only the results for the exper-

iments showing a pH higher than 2, in any of the bottles, are shown in the table.

Table 5.3: pH measured in the samples from the gas washing bottles, flask 2 is in parentheses.

M1 M2 M3 M4

T (°C) 75 75 65 75

O₂ (%) 98 6 50 50

Sample 1 9-10 (0) 0-1 (0) 9 (0) 9 (0)

Sample 2 9-10 (8-9) 1-2 (0) 8-9 (0) 1-2 (0)

Sample 3 1 (0) 0-1 (0) 2 (0) 8-9 (0)

Sample 4 9-10 (0) 1 (0) 1 (0) 3-4 (0)

Sample 5 9-10 (0) 0-1 (0)

Sample 6 1 (0)

Sample 7 0-1 (0)

The results show that only sample two of experiment M1 consumed all the amount of acid in both

bottles. This means that all the ammonia in the gas phase was not captured for this experiment. The

results for ammonia in the gas washing bottles of M1 are therefore assumed to show less ammonia

than what was released as gas.


31

5.2. Overall degradation in experiments

5.2.1. Nitrogen balance

The total weight percent of nitrogen in the end samples analyzed by the Kjeldahl method [52], was

compared to the weight percent of nitrogen containing compounds accounted for from LC-MS and

NH3 analyses. The results for the project and master thesis are presented in Table 5.4.

Table 5.4: Nitrogen recovery from end samples.

Experiment T (°C) O₂ (%) Time

(Days)

Kjeldahl

(wt%) MEA (%)

N in known com-

pounds (wt%)

Nitrogen recovery

(RN, %)

M5 55 6 42 5.93 4.82 0.08 83

P2 55 21 21 6.70 5.44 0.18 84

V1 55 21 21 6.00 5.34 0.24 93

V2 55 21 21 5.70 5.36 0.11 96

V3 55 50 21 5.80 5.16 0.17 92

V4 55 98 21 5.60 4.59 0.22 86

M6 65 6 42 5.76 5.28 0.12 94

P4 65 21 21 5.93 4.76 0.18 83

M3 65 50 28 5.38 3.05 0.38 64

P3 65 98 21 5.37 2.99 0.43 64

M2 75 6 36 4.63 4.52 0.21 102

V5 75 21 21 4.60 2.59 0.28 62

M4 75 50 28 4.77 2.07 0.35 51

M1 75 98 21 5.17 1.99 0.43 47

The MEA results for experiment M5 and M6 have an uncertainty of up to ±10%, see Section 4.1. The

experiments performed by Vevelstad have an uncertainty of up to ±5%. The trends in the results

indicate that the oxygen concentration does not influence the amount of nitrogen containing com-

pounds for the low temperature experiments. For the experiments at 65 and 75 °C, the nitrogen re-

covery is low when the oxygen concentration is more than 6%. The temperature does not seem to

influence the amount of unknown degradation compounds, when the oxygen concentration is low.

This indicates that the unknown degradation compounds formed are both temperature and oxygen

dependent.

In

Table 5.5, the molar balance for nitrogen in MEA, Equation 5.1, is shown. The formation of each of

the degradation compounds were calculated as shown by Lepaumier [50], by the Equation 5.2. ,f i is

the formation percentage for each degradation compound and i is the number of nitrogen atoms in

each compound. o

MEAC ,end

MEAC and end

iC is the molar concentration of MEA in the start and end sam-

ple, and of a degradation compound i in the end sample, respectively. N unidentified is the for-

mation percentage of unidentified compounds from the initial concentration of MEA.


32

o end end

MEA MEA i

i

C C C (5.1)

, 100end

if i i o

MEA

C

C (5.2)

Table 5.5: Molar balance for nitrogen in MEA, and formation percentage of all degradation compounds.

M5 P2 M6 P4 M3 P3 M2 M4 M1

T (°C) 55 55 65 65 65 65 75 75 75

O₂ (%) 6 21 6 21 50 98 6 50 98

Time (Days) 42 21 42 21 28 21 21 28 36

MEA start (mol/L) 5.71 5.22 5.53 5.22 5.26 5.25 4.62 5.43 5.27

MEA end (mol/L) 4.52 4.28 4.46 3.72 2.36 2.31 3.47 1.58 1.52

MEA loss (%) 21 18 19 29 55 56 25 71 71

% formation

DEA 0.0053 0.0084 0.0061 0.0075 0.0070 0.0055 0.012 0.004 0.010

NH₃ 0.16 0.50 0.25 0.87 2.1 2.3 0.49 1.6 1.6

HeGly 0.27 0.50 0.62 0.34 0.21 0.14 1.3 0.20 0.30

HEF 0.52 1.28 0.53 0.89 1.98 2.15 0.87 1.66 2.19

BHEOX 0.046 0.056 0.035 0.079 0.15 0.18 0.036 0.093 0.091

HEA 0.030 0.060 0.059 0.047 0.15 0.094 0.10 0.27 0.22

HEPO 0.046 0.078 0.086 0.067 0.08 0.059 0.12 0.079 0.15

OZD 0.082 0.079 0.050 0.093 0.42 0.30 0.041 0.54 0.26

HEI 0.10 0.49 0.28 0.68 1.29 2.33 0.99 1.34 3.20

Total organic 1.25 3.05 1.92 3.08 6.35 7.58 3.99 5.74 8.00

NO2 0.024* 0.192 0.033* 0.162 0.90 0.78 0.02* 0.14* 1.39

NO3 0.068* 0.043 0.055* 0.056 0.30 0.37 0.03* 0.64* 0.57

Total inorganic 0.092* 0.235 0.088* 0.218 1.201 1.152 0.05* 0.78* 1.967

NH₃ from flasks 2.68 - 4.37 - 9.54 10.42 7.61 12.87 12.14

Total MEA loss

from degradation

compounds (%) 4 3 6 3 17 19 12 19 22

N unidentified (%) 17 15 13 25 38 37 13 52 49 *Results were automatically generated from IC-EC, and are very uncertain.

The formation of NH3, HEF and HEI seems to be the main contributors to the degradation of MEA.

This is in accordance with the results found by Vevelstad [37]. The NH3 values from the flasks are

recalculated from the concentrations in the gas washing bottle, and due to the loss of acidity for both

washing bottles in experiment M1, the NH3 value for the flasks in M1 is uncertain. For experiment

M2, NH3 is accounting for about a third of the degradation of MEA and is by far the main degrada-

tion compound.


33

5.2.2. Kinetics

The reaction rate of the experiments was investigated by calculating the rate constant for all the

experiments with regards to MEA.

The rate expression for oxidative degradation of MEA as described by Supap et al, is shown in Equa-

tion 5.3, where r is the reaction rate, k is the rate constant and [MEA] is the MEA concentration.

[ ]dC

r k MEAdt

(5.3)

The reaction of MEA to the various degradation compounds is assumed to be first order with regards

to MEA. Integrating the rate expression over the time t and rearranging the expression gives Equa-

tion 5.4.

[ ]ln

[ ]

oMEAk t

MEA (5.4)

Plotting ln([MEA]°/[MEA]) against t gives a slope of k. The rate constants for the experiments were

also calculated for second order reactions by plotting 1/[MEA] versus t. The plots for experiment M3

are shown in Figure 5.1.

Figure 5.1: Plot for first (a) and second (b) order reaction for experiment M3. The rate constant k is the value

found in front of x in the equation.

The coefficients of determination, R2, were very similar for both the first order and second order

plots. This can indicate that MEA goes through a second order reaction rather than a first order reac-

tion. This is although quite unlikely as this would mean that MEA polymerizes in the reaction vessel.

Therefore the rate constants for a first order reaction with regards to MEA are compared.

The rate constants k, were found for all experiments by plotting. The values obtained are shown in

Figure 5.2.


34

Figure 5.2: Rate constants at different oxygen concentrations and temperatures for all experiments. Square markers are the experiments performed by Vevelstad[37].

The plot shows that the reaction rate for the experiments is not greatly affected by the oxygen flow

when the temperature is low. The reaction rate increases however by a factor 3 going from 65°C to

75°C at 98% oxygen. This indicates that some of the degradation reactions take place at a tempera-

ture above 65°C.The first order reaction rates found for the experiments are listed in Table 5.6.

Table 5.6: Rate constant k for all experiments

Experiment O2 (%) T (C) k

M5 6 55 °C 0.0061

P2 21 55 °C 0.0096

V1 21 55 °C 0.0083

V2 21 55 °C 0.0074

V3 50 55 °C 0.0088

V4 98 55 °C 0.014

M6 6 65 °C 0.0069

P4 21 65 °C 0.0115

M3 50 65 °C 0.0273

P3 98 65 °C 0.0349

M2 6 75 °C 0.0121

V5 21 75 °C 0.0454

M4 50 75 °C 0.0424

M1 98 75 °C 0.1067

5.2.3. Amine loss

The amine concentration is calculated from the titration results for the different experiments dis-

cussed. The amine loss in percent, α, is calculated by Equation 5.5. o

mC and i

mC are the molar concen-

trations of amines in the start sample and in sample I, respectively. For these calculations sample I is

the end sample. The amine loss for the experiments is shown in

Figure 5.3.


35

100o i

m m

o

m

C Ca

C

(5.5)

Figure 5.3: Amine loss (%) for experiments performed with different oxygen concentrations and temperatures.

As can be seen from the figure, the amine loss is significantly higher for the experiments at 65 and 75

°C compared to the 55 °C experiments. An increase in the oxygen concentration of the volumetric gas

flow does not seem to increase the amine loss after it has reached 50%.

The development of the amine concentration for the experiments performed at 75 °C, and the exper-

iments with an oxygen concentration of 6% are shown in Figure 5.. The normalized amine concentra-

tion is found using Equation 5.6.

i

mN o

m

CC

C (5.6)

(a) (b)

Figure 5.4: Reduction in amine concentration over time for experiments at 75 °C (a), and experiments at 6%

oxygen concentration.


36

5.2.4. Loading of CO2 in end samples

The loading of CO2 in MEA (α) was calculated by dividing the molar concentration of CO2 with the

amine concentration for all the end samples of the experiments. The measured values for the CO2

concentrations can be found in Appendix D. The results are illustrated in Figure 5.5.

Figure 5.5: Loading of end samples for experiments at different temperatures and oxygen concentrations. The

square points are the experiments performed by Vevelstad [37].

The initial loading α, of the samples was 0.4 and for some of the experiments the loading increased.

This might be due to differences in the calibration of the gas flow as discussed in chapter 4. The load-

ing does not change significantly for the 55 and 65 °C experiments. For the 75 °C experiments the

loading decreases with increasing oxygen concentration. Some of the decrease in loading might be

caused by desorption of CO2 which happens when the temperature rises, but this is generally for

temperatures over 121 °C [53]. A decrease in loading might suggest the formation of thermal degra-

dation compounds. Thermal degradation compounds are found to be formed when CO2 is present,

suggesting that the formation mechanism involves CO2[56]. It has been shown [57] that the loading

has a first order effect on the degradation rate of MEA, and therefore this is believed to have an im-

pact on the speed of degradation. Assuming that the loading has decreased linearly, the greatest

effect will be on the last samples of the experiments.

5.2.5. Anion balance

The total concentration of anions in the end samples were compared with the results for heat-stable

salts (HSS). The results are given in


37

Table 5.7. From the anions formate contributes to 75% of the anions analyzed in all the samples. Oxa-

late and nitrite have about the same degradation contributing to around 10% each, and nitrate had a

contribution of 4%.


38

Table 5.7: Concentration of HSS compared to analyzed anions in end samples.

T (°C) O₂ (%) Time (Days) HSS (eq/kg) Total identified ani-

ons (mol/kg) Identified (%)

M5 55 6 42 0.089 0.05 59

P2 55 21 21 0.264 0.07 26

M6 65 6 42 0.138 0.09 62

P4 65 21 21 0.248 0.13 54

M3 65 50 28 0.595 0.32 53

P3 65 98 21 0.650 0.37 57

M2 75 6 36 0.230 0.18 80

M4 75 50 28 0.835 0.61 73

M1 75 98 21 1.009 0.75 74

The trend show that HSS, as the other degradation compounds, are formed at higher temperatures

and oxygen concentrations. The unidentified HSS might be due to the presence of amides, as shown

by Vevelstad[37]. It does not seem to be a distinct trend between the amount of HSS formed in an

experiment and the amount of HSS accounted for by anion analysis.


39

5.3. Degradation compounds

In this section the results for the first and second order degradation compounds will be presented

and discussed with regards to the proposed mechanisms given in Section 2.2.

5.3.1. First order degradation compounds

The first order degradation compounds were analyzed by the IC-EC method. As discussed in chapter

4, this method showed some inconsistency in the quantification, and will therefore give some uncer-

tainties in the results. The concentrations (PPM) found for all the analyzed anions are given in Ap-

pendix D.

Formate

The formate was the first order degradation compound that was found to have the highest concen-

tration of anions in the degradation solution. A graph showing the concentration of formate in

mmol/kg after 21 days for all experiments is shown in Figure 5.6.

Figure 5.6: Concentration of formate for samples at 21 days at different reaction temperatures and oxygen percentages.

The trends in the figure seem to indicate that the development of formate is influenced highly by the

temperature of the reactor. For the experiments performed at 98% oxygen the formate concentra-

tion after 21 days is about 7 times as high for the experiment performed at 75 °C compared to the

experiment at 55 °C. The concentration of oxygen in the gas stream seems to influence the formation

of formate in a linear way until the oxygen concentration reaches 50%. This supports the literature

stating that formation of formate involves oxygen. From 50 to 98% the oxygen does not influence the

formation of formate in a significant way.

The development of formate throughout the experiment is shown for some experiments in Figure

5.7.


40

Figure 5.7: The development of the formate concentration as a function of time for some of the performed experiments.

The figure shows a trend where the formation of formate gradually slows down for experiment M1

and M3. This might be due to the high degradation of the reactant MEA, but it can also indicate that

formate is consumed to produce secondary degradation compounds (SDC). Formate is as stated in

Section 2.2 believed to participate in the formation of HEF.

Oxalate

The concentration of the oxalate anion was found to be at about a tenth of the concentration of

formate for the experiments combined. The oxalate concentrations found in mmol/kg for the exper-

iments after 21 days is shown in Figure 5.8.

Figure 5.8: Concentration of oxalate for samples at 21 days at different reaction temperatures and oxygen percentages.

The trends seen in the accumulated concentrations of oxalate under different conditions seem to be

similar to the trends seen for formate. This also supports the statement that both anions are formed

through a similar mechanism involving oxygen.


41

Looking at the development of oxalate over time for some of the experiments, the anion seems to

growth proportionally over time for the experiments performed at high oxygen percentages and

temperatures. This is shown in Figure 5.9.

Figure 5.9: The development of the oxalate concentration as a function of time for some of the performed experiments.

For the experiments performed with a low oxygen concentration the oxalate seems to have a more

exponential growth. This might be due to oxalate forming from some of the other degradation com-

pounds.

Nitrate

Nitrate was found in the smallest concentration of the anions analyzed for. It was at about half the

concentration of oxalate and 20 times smaller than formate. The concentration of nitrate found in

the solution after 21 days is shown in Figure 5.10.

Figure 5.10: Concentration of nitrate for samples at 21 days at different reaction temperatures and oxygen percentages. Point at 21% O2 and 75 °C is from experiment by Vevelstad.

The concentration of nitrate found in the solution seems to increase when it comes to the oxygen

concentration in the gas stream. This is also the case for the temperatures at the highest oxygen con-


42

centration. At the lower oxygen concentrations the trend also indicates that nitrate increases with

the temperature, but this is difficult to say anything certain about due to the variation in the data.

How nitrate has evolved in the reactors for some of the experiments is shown in Figure 5.11. The

uncertainty in the values is up to 35%. This might be due to difficulties in quantifying compounds of

low concentration.

Figure 5.11: The development of the nitrate concentration as a function of time for some of the performed experiments.

From the graph it seems like nitrate also has a linear development over time. The concentration of

nitrate in the samples seems to double when the oxygen concentration doubles. If a better method is

found for analyzing anions by IC-EC it would be possible to see if the formation of nitrate gradually

slows down for M1.

Nitrite

The concentration of nitrite found in the experiments is at about the same level as the oxalate con-

centration. The nitrite sample concentration in the experiments after 21 days is shown in Figure 5.12.

Figure 5.12: Concentration of nitrate for samples at 21 days at different reaction temperatures and oxygen percentages.


43

Nitrite also seems to show a linear trend of increase in concentration relative to the oxygen flow. The

high concentration of nitrite for 50% O2, 65 °C, is most likely due to uncertainties in the analytical

method. The development of nitrite shows a similar degradation pattern as nitrate

Figure 5.13: The development of the nitrite concentration as a function of time for some of the performed experiments.

In Figure 5.13 the development of nitrite over time for some of the experiments is shown. In almost

all the experiments the growth of nitrite slows down after about 21 days. For experiment M1 the

amount of nitrite in the reactor decreases after a peak at about 10 days. Nitrate and nitrite is be-

lieved to be formed though oxidation of NOx formed from oxidized nitrogen [38]. The decrease in

nitrate for M1 can either be caused by the nitrite reacting on to other degradation compounds, or by

the anions being reduced back to nitrogen oxides.

5.3.2. Secondary degradation compounds (SDC)

HEF

HEF is the SDC analyzed for by LC-MS found in the highest concentration in total for the experiments.

The amount of HEF found in the samples taken after 21 days for all the experiments are shown in

Figure 5.14.

Figure 5.14: Concentration of HEF in samples at 21 days at different reaction temperatures and oxygen percentages.


44

Figure 5.14 show that the development rate of HEF does not change significantly when the oxygen

concentration goes from 50% to 98%. HEF is believed to be formed through the mechanism shown

in Figure 2.6, where formate is believed to participate in the formation. When comparing the results

with the trends seen for formate, the formation of HEF does not increase in the same rate as formate

with regards to the oxygen concentration. This can be due to HEF being involved in reactions that

take place at a higher oxygen concentration, and therefore is consumed. This also explains why the

HEF concentration for 75 °C is larger at 6% oxygen compared to 21% oxygen. The temperature does

not seem to affect the development of HEF in any significant way.

In Figure 5.15 the development of HEF over time is compared for some of the experiments.

Figure 5.15: The development of the HEF concentration as a function of time for some of the performed experiments.

The results show that for M1 the HEF concentration has a peak at about 10 days. This is the same

trends as is seen for nitrite. This can be caused by HEF and nitrite needing similar conditions for for-

mation, but it can also be a result of the two compounds being involved in the formation of another

degradation compound. To investigate if this is the case, a mixture experiment of HEF and nitrite can

be performed, and analyzing the solution for new products. The formation of HEF seems to be de-

pending on the oxygen flow when the oxygen concentration and temperature is low. For high oxygen

concentrations the formation seems however more temperature dependent.

HEA

HEA is the third smallest degradation compound analyzed for in the degradation mix, and has a for-

mation of about a tenth of HEF. The formation of HEA after 21 days is shown for the experiments in

Figure 5.16.


45

Figure 5.16: Concentration of HEA for samples at 21 days at different reaction temperatures and oxygen percentages.

The trend shows that HEA is formed at a higher rate with increasing temperatures. HEA is assumed to

be formed from acetic acid, which had not been analyzed because of low detection limits, and the

trends shown in the graph cannot be compared to see if the development is similar. As can be seen

from the graph the formation of HEA has a peak at an oxygen concentration of 50%. This can be due

to either less acetic acid in the high temperature experiments, or as a result of acetic acid reacting to

other compound at these conditions. It can also be explained by HEA being an intermediate in the

formation of other degradation compounds at high temperatures. As seen in Figure 5.17, HEA seems

to increase about linearly over time, but for experiment M1, the concentration decreases slightly

over the last week.

Figure 5.17: The development of the HEF concentration as a function of time for some of the performed experiments.

BHEOX

BHEOX accounts for 2% of the combined results found for SDCs in the experiments after 21 days. The

concentration of BHEOX found in the solution after 21 days is shown in Figure 5.18.


46

Figure 5.18: Concentration of BHEOX for samples at 21 days at different reaction temperatures and oxygen percentages.

The trends seen for the concentration of BHEOX at 21 days show that the experiments performed at

the highest temperature has the smallest concentration. The experiments at 55 and 65 °C seem to be

formed at about the same rate. As for HEF the oxygen concentration has the most pronounced effect

between 6 and 50%.

Comparing the results of the experiments over time gives the graphs shown in Figure 5.19. It seems

like the amount of BHEOX decreases at the end for almost all the experiments. The mechanism of

formation of BHEOX is known to be a reversible mechanism with oxalate. As seen in Section 5.3.1 the

formation of oxalate increases at high temperatures. It seems that the formation of oxalate is fa-

vored over time, which is also supported by its formation rate over time seen for the experiments

with low oxygen concentration. The decrease in concentration for BHEOX over time, especially for

the high temperature experiments, might also be caused by other degradation compounds formed at

high temperatures with BHEOX as an intermediate.

Figure 5.19: The development of the BHEOX concentration as a function of time for some of the performed experiments.

OZD

The OZD concentrations contributed to about 10% of the SDCs found by analysis. The formation of

OZD relative to the oxygen concentration seems to be pronouncedly favored for a 50% concentration

of oxygen in the gas stream, as shown in Figure 5.20. This is seen especially for the experiments per-

formed at 65 and 75 °C.


47

Figure 5.20: Concentration of OZD for samples at 21 days at different reaction temperatures and oxygen percentages.

The mechanism for formation of OZD is suggested to be induced by CO2. The concentration of CO2 in

the solution is significantly reduced at the experiments with high oxygen concentration and tempera-

ture, and may be the cause of the low formation of OZD at these points. The increase in the for-

mation of OZD at 50% oxygen might also be explained by further reactions of OZD at higher oxygen

percentages. If this is the case the mechanism for formation most likely involves oxygen.

Figure 5.21: The development of the OZD concentration as a function of time for some of the performed experiments.

Figure 5.21 shows the development of the OZD concentration in the experiments over time. From

the graphs it seems like the formation rate is similar for the experiments with oxygen concentration

over 50% at a temperature of 65 degrees or higher. Experiment M1 shows a slight reduction in OZD

at the end of the experiment.

HEI

The concentration of HEI found for the experiments after 21 days corresponds to about 20% of the

total amount of SDCs analyzed for. The formation of HEI seems to increase linearly with regards to

the oxygen concentration. The exception is the experiment at 75 °C and 6% oxygen (M2), as seen in

Figure 5.22, which shows the same trend as seen for HEF. This can indicate a relation between the

mechanisms for HEI and HEF. If this is due to HEI being an intermediate for further reaction, the reac-

tion that takes place is both depending on temperature and oxygen concentration.


48

Figure 5.22: Concentration of HEI for samples at 21 days at different reaction temperatures and oxygen percentages.

It is difficult to say anything about the formation of HEI from the assumed intermediated, since the

samples were not analyzed for the aldehydes believed to be involved. It is although likely the HEI is

formed from through multiple reactions, since the formation is both depending on oxygen and tem-

perature. The concentration of HEI at 55 °C seems to increase linearly with time, but for the experi-

ments performed at higher temperatures the formation slows down over time. In experiment M1,

HEI decreases at the end of the experiment, indicating further degradation.

Figure 5.23: The development of the HEI concentration as a function of time for some of the performed experiments.

HEPO

HEPO is the SDC that has been found in the smallest concentration for the total group of experi-

ments. In contrast to the other SDCs HEPO has a decrease of formation at 50% oxygen for the 75°C

experiments, as shown in figure 5.24. In figure 5.25 the development of HEPO for this experiment is

shown. It seems like the values given for the sample concentrations of this experiment do not follow

a smooth curve, and it might therefore be some uncertainty in the value given. The values will also

be more uncertain because of difficulties analyzing for very small concentrations.


49

Figure 5.24: Concentration of HEPO for samples at 21 days at different reaction temperatures and oxygen percentages.

If the discussed point is ignored, the concentration of HEPO seems to not depend on the oxygen con-

centration. The concentration does however seem to have a linear relationship with the temperature

for most of the experiments. The curves for the concentration over time show that, for the experi-

ments with high degradation, the rate of formation decreases towards the end of the experiments.

For the low degradation experiments, the formation rate seems to grow over time.

Figure 5.25: The development of the HEPO concentration as a function of time for some of the performed experiments.

HEGly

HEGly is a secondary degradation product that is contributes to about 12% of the total concentration

of the analyzed SDCs. The formation of HEGly in the performed experiments after 21 days is shown in

Figure 5.26.


50

Figure 5.26: Concentration of HEGLy for samples at 21 days at different reaction temperatures and oxygen percentages.

From the figure it is evident that HEGly is formed at a much higher rate when the oxygen concentra-

tion is low (6%). For the 6% experiments the formation also seems to have a linear relationship with

temperature, whereas the experiments at 21% and higher does not seem to show a very distinct

temperature dependency.

The trends seen in the development of HEGly over time is that the formation rate of HEGly decreases

over time. This is shown in Figure 5.27. For experiment M4(75 °C, 50% O2), the concentration of

HEGly decreases after about 7 days. This can indicate that HEGly is part of an unknown reaction lead-

ing to other unknown degradation compounds. This might also mean the formation of HEGly is in a

relationship with OZD or HEA, which show a favored formation at 50% oxygen.

The formation mechanism of HEGly has been important to understand as HEGly was found to be

among the dominant degradation compounds in MEA pilot plant samples [32]. A new mechanism for

HEGly has been suggested, and mixing experiments have been performed to investigate its validity.

This is discussed in Section 5.4.

Figure 5.27: The development of the HEGLy concentration as a function of time for some of the performed experiments.


51

5.4. Mixing experiments for HEGly

A credible formation mechanism for HEGly has not yet been verified. Previously assumed precursor

such as DEA and glycine, which both show a similar structure to HEGly, had not seemed to produce

significant amounts of HEGly when mixed [58]. A new mechanism of formation has therefore been

proposed where glycol aldehyde or glyoxylic acid has been assumed to be the precursors, see Figure

5.28.

Figure 5.28: Proposed mechanism for the formation of HEGly

The proton transfer steps are however uncertain, and may be affected by pH, components in the

solution and reaction conditions. Molecular modelling can be used to suggest which proton transfer

mechanisms are more likely to occur. Eide-Haugmo has shown how to do quantum mechanical calcu-

lations to find the reaction energy for various reactions[56]. This can show which mechanical step is

more favorable for reaction.

To investigate if the suggested mechanism is plausible two mixing experiments were performed by

mixing glyoxylic acid, formic acid and MEA. The first reaction solution was acidic and prepared by

having an excess amount of formic acid, and MEA as the limiting component. The second experiment

was run with glycolic acid as the limiting component and an excess of MEA, making the solution basic

and more similar to the actual conditions of the CO2 absorption column. The concentration of the

degradation compound HEGly found for the experiments are shown in Figure 5.8. The values used for

the valculation is given in Appendix E.

Figure 5.8: Formation of HEGly from glyoxylic acid (molar%) for mixing experiments over time.


52

The results show that HEGly was formed at a high rate for experiment 1. The formation of HEGly

reached 10% of the original molar concentration of glyoxylic acid after 3 days. This result seems to

support the suggested mechanism. For experiment 2 the formation of HEGly was only 0.14% of the

original concentration of glyoxylic acid. HEF was however formed from 17% of the glyoxylic acid con-

centration. This might be a result of HEF forming from formic acid and MEA, but HEF was also found

in the solution when only glyoxylic acid had been added.

The results found from the second mixing experiment seem to contradict the suggested mechanism,

but this may also be because the formation of HEGly is slower in basic conditions. Because the exper-

iment only was run for 2 days, more experiments are needed to fully understand if the mechanism is

plausible or not.

53

6. Conclusion

For more comparable results of the experimental values, the mass flow controllers should be cali-

brated regularly and the water balance should be monitored by the amount of water in the wetting

chambers.

The nitrogen balance shows that changing the temperature does not seem to influence the amount

of nitrogen containing degradation compounds with low oxygen concentration (6%) in the gas

stream. For experiments performed at 65 degrees or higher, the formation of unknown degradation

compounds increases when the oxygen concentration goes from low (6%) to medium (50%), but

does not seem to be affected significantly when the oxygen concentration is increased beyond that.

The reaction rate of the degradation of MEA show the same trends that are seen in the nitrogen bal-

ance. The development seen of the MEA concentrations is in accordance with the degradation reac-

tion being first order with regards to MEA. The degradation would also fit the curve for a second

order reaction of MEA, but this is highly unlikely. The amine loss for the experiments show the same

trends for degradation as seen in the nitrogen balance.

The loading of CO2 to the amine concentration seems not to be close to constant for the experiments

performed at 55 and 65 °C. At 75°C the loading decreases when the oxygen flow is increased. This

suggests that there is some formation of thermal degradation compounds at 75 °C.

First order degradation compounds found in the solution as anions are found to be present in the

solutions in the order: formate > oxalate/nitrite > nitrate. The dependency on oxygen concentration

and temperature for the anions are given in Table 6.1.

Table 6.1: Trends seen for anions relative to oxygen concentration and temperature. Yes means the trend is seen in all

experiments

Anion Oxygen dependent Temperature dependent

Formate Up to 50% O2 Yes

Oxalate 55-65 °C: Up to 50% O2, 75°C:Yes Yes

Nitrate Yes 98% O2:Yes, other exp: Unknown

Nitrite Yes 98% O2:Yes, other exp: Unknown

The formation of compounds seen in the total mass of experiments, give the following order for sec-

ondary degradation products: HEF > HEI > HEGly > OZD > HEA > BHEOX > HEPO. The trends for these

degradation compounds regarding oxygen concentration and temperature is given in Table 6.2.

54

Table 6.2: Trends seen for degradation compounds relative to oxygen concentration and temperature. Yes means the

trend is seen in all experiments

Secondary degradation compound Oxygen dependent Temperature dependent

HEF From 6 to 50% In early stages for high O2(%)

HEI Yes Yes, mainly 55-65 °C

HEGly Favored at 6% Mainly for 6% O2

OZD Peak at 50% 6% to 50% O2: Yes

HEA Peak at 50% Yes

BHEOX Mainly from 6 to 50% Less formed at 75 °C

HEPO No Yes

The mechanism for the formation of HEGly suggested by Vevelstad seems to be unlikely, but further

experiments are needed to conclude.

55

7. Suggestions for further work

There is still more work needed to be done to improve the understanding of the degradation of MEA.

For the oxidative degradation setup, running experiments over a longer period of time would in-

crease the understanding of the development of some of the degradation compounds. It would be of

interest to rerun some of the experiments to see if the results are reproducible. This is mainly for the

experiments run at high temperatures and oxygen percentages as these experiments often show

more variation in the curves of formation.

For the experiments that has been performed during this work, more accurate methods for quantifi-

cation is needed to be developed for both LC-MS, GS-MS and IC-EC. The MEA quantification by LC-MS

should be altered since the given concentrations does not seem to be in accordance with the calcu-

lated values. The quantification by GC-MS seemed to work better using methanol as a solvent, and

higher dilutions of the samples may give better results. Finding ways to analyze and quantify the

development of more of the suggested degradation compounds will give a better picture of which

compounds are formed from what precursors.

To investigate which unidentified degradation compounds are formed, synthesis experiments mixing

known degradation compounds can be conducted. It would be interesting to analyse the reaction

mixture after separating the solution using column chromatography and analyzing the different com-

pounds by NMR. Mixing nitrite and BHEOX would be interesting in order to investigate if they form

any known or unidentified degradation compounds, as described in Section 5.3.2. If unidentified deg-

radation compounds are found, mechanisms of formation should be suggested.

To find more evidence for the formation mechanism of HEGly or to dismiss it, more experiments

similar to the mixing experiments performed are needed. Some suggested reactions are mixing glycol

aldehyde with formic acid and MEA. These experiments should also be performed over a larger time

interval, compared to the mixing experiments performed for this thesis, to get a more detailed un-

derstanding of the reaction.

56

References

1. Pachauri, R.K., Climate change 2007. Synthesis report. Contribution of Working Groups I, II and III to the fourth assessment report. 2008.

2. Meinshausen, M., et al., Greenhouse-gas emission targets for limiting global warming to 2[thinsp][deg]C. Nature, 2009. 458(7242): p. 1158-1162.

3. Van Aalst, M.K., The impacts of climate change on the risk of natural disasters. Disasters, 2006. 30(1): p. 5-18.

4. Rodhe, H., A comparison of the contribution of various gases to the greenhouse effect. Science, 1990. 248(4960): p. 1217-1219.

5. Keeling, C.D., Rewards and penalties of monitoring the Earth. Annual Review of Energy and the Environment, 1998. 23(1): p. 25-82.

6. Metz, B., et al., Carbon dioxide capture and storage. 2005. 7. Metz, B., et al., IPCC special report on carbon dioxide capture and storage: Prepared by

working group III of the intergovernmental panel on climate change. IPCC, Cambridge University Press: Cambridge, United Kingdom and New York, USA, 2005. 2.

8. Buhre, B.J.P., et al., Oxy-fuel combustion technology for coal-fired power generation. Progress in Energy and Combustion Science, 2005. 31(4): p. 283-307.

9. Lyngfelt, A., B. Leckner, and T. Mattisson, A fluidized-bed combustion process with inherent CO2 separation; application of chemical-looping combustion. Chemical Engineering Science, 2001. 56(10): p. 3101-3113.

10. Gibbins, J. and H. Chalmers, Carbon capture and storage. Energy Policy, 2008. 36(12): p. 4317-4322.

11. Rochelle, G.T., Amine scrubbing for CO2 capture. Science, 2009. 325(5948): p. 1652-1654. 12. Notz, R., et al., Selection and Pilot Plant Tests of New Absorbents for Post-Combustion Carbon

Dioxide Capture. Chemical Engineering Research and Design, 2007. 85(4): p. 510-515. 13. Rao, A.B. and E.S. Rubin, A Technical, Economic, and Environmental Assessment of Amine-

Based CO2 Capture Technology for Power Plant Greenhouse Gas Control. Environmental Science & Technology, 2002. 36(20): p. 4467-4475.

14. Abu-Zahra, M.R.M., et al., CO2 capture from power plants. Part I. A parametric study of the technical performance based on monoethanolamine. International Journal of Greenhouse Gas Control, 2007. 1(1): p. 37-46.

15. Bedell, S.A., Amine autoxidation in flue gas CO2 capture—Mechanistic lessons learned from other gas treating processes. International Journal of Greenhouse Gas Control, 2011. 5(1): p. 1-6.

16. Bedell, S.A., Oxidative degradation mechanisms for amines in flue gas capture. Energy Procedia, 2009. 1(1): p. 771-778.

17. Kennard, M.L. and A. Meisen, Mechanisms and kinetics of diethanolamine degradation. Industrial & engineering chemistry fundamentals, 1985. 24(2): p. 129-140.

18. Kim, C., Degradation of alkanolamines in gas-treating solutions: kinetics of di-2-propanolamine degradation in aqueous solutions containing carbon dioxide. Industrial & engineering chemistry research, 1988. 27(1): p. 1-3.

19. Chi, S. and G.T. Rochelle, Oxidative degradation of monoethanolamine. Industrial & engineering chemistry research, 2002. 41(17): p. 4178-4186.

20. Dennis Jr, W.H., L.A. Hull, and D.H. Rosenblatt, Oxidations of amines. IV. Oxidative fragmentation. The Journal of Organic Chemistry, 1967. 32(12): p. 3783-3787.

21. Rosenblatt, D., et al., Oxidations of amines. II. Substituent effects in chlorine dioxide oxidations. Journal of the American Chemical Society, 1967. 89(5): p. 1158-1163.

57

22. Beckwith, A.L., et al., Amine autoxidation in aqueous solution. Australian Journal of Chemistry, 1983. 36(4): p. 719-739.

23. Audeh, C. and J.L. Smith, Amine oxidation. Part II. The oxidation of some trialkylamines with alkaline potassium hexacyanoferrate (III). J. Chem. Soc. B, 1970: p. 1280-1285.

24. Smith, J.L. and L. Mead, Amine oxidation. Part VII. The effect of structure on the reactivity of alkyl tertiary amines towards alkaline potassium hexacyanoferrate (III). J. Chem. Soc., Perkin Trans. 2, 1973(2): p. 206-210.

25. Smith, J.L. and L. Mead, Amine Oxidation. Part IX. Oxidation of Some Substituted Tertiary Alkylamines and Some N, N-dimethylphenethylamine with Potassium Hexacyanoferrate (III). J. Chem. Soc. Perkin II, 1976: p. 1172-1176.

26. Rosenblatt, D.H., et al., Oxidations of amines. V. Duality of mechanism in the reactions of aliphatic amines with permanganate. The Journal of Organic Chemistry, 1968. 33(4): p. 1649-1650.

27. Petryaev, E., A. Pavlov, and O. Shadyro, Homolytic deamination of amino alcohols. Zh. Org. Khim, 1984. 20(1): p. 29-34.

28. Alejandre, J., et al., Force field of monoethanolamine. The Journal of Physical Chemistry B, 2000. 104(6): p. 1332-1337.

29. Button, J., et al., Molecular dynamics simulation of hydrogen bonding in monoethanolamine. Fluid phase equilibria, 1996. 116(1): p. 320-325.

30. Yazvikova, N., L. Zelenskaya, and L. Balyasnikova, Mechanism of side reactions during removal of carbon dioxide from gases by treatment with monoethanolamine. Zhurnal Prikladnoi Khimii, 1975. 48(3): p. 674-676.

31. Gouedard, C., et al., Amine degradation in CO2 capture. I. A review. International Journal of Greenhouse Gas Control, 2012. 10: p. 244-270.

32. da Silva, E.F., et al., Understanding 2-Ethanolamine Degradation in Postcombustion CO2 Capture. Industrial & Engineering Chemistry Research, 2012. 51(41): p. 13329-13338.

33. Lepaumier, H., et al., Comparison of MEA degradation in pilot-scale with lab-scale experiments. Energy Procedia, 2011. 4: p. 1652-1659.

34. Katsuura, A. and N. Washio, Preparation of imidazoles from imines and iminoacetaldehydes. 2005, Nippon Synthetic Chemical Industry Co., Ltd., Japan . p. 6 pp.

35. Kawasaki, N., et al., Preparation of 1-substituted imidazoles. 1991, Mitsui Toatsu Chemicals, Inc., Japan . p. 7 pp.

36. Ben, D.S.P., Process for the preparation of -1-(2-hydroxyethyl) imidazole. 2005, India . p. 6pp. 37. Vevelstad, S., CO2 absorbent degradation, in Department of Chemical Engineering. 2013,

Norwegian University of Science and Technology: Trondheim. 38. Sexton, A.J. and G.T. Rochelle, Reaction products from the oxidative degradation of

monoethanolamine. Industrial & Engineering Chemistry Research, 2010. 50(2): p. 667-673. 39. Supap, T., et al., Analysis of monoethanolamine and its oxidative degradation products during

CO2 absorption from flue gases: A comparative study of GC-MS, HPLC-RID, and CE-DAD analytical techniques and possible optimum combinations. Industrial & engineering chemistry research, 2006. 45(8): p. 2437-2451.

40. Kadnar, R. and J. Rieder, Determination of anions in amine solutions for sour gas treatment. Journal of Chromatography A, 1995. 706(1): p. 339-343.

41. Christie, W.W., Gas chromatography and lipids. 1989. 42. Poppe, H., Some reflections on speed and efficiency of modern chromatographic methods.

Journal of Chromatography A, 1997. 778(1): p. 3-21. 43. Greibrokk, T. and T. Andersen, High-temperature liquid chromatography. Journal of

Chromatography A, 2003. 1000(1): p. 743-755. 44. Strazisar, B.R., R.R. Anderson, and C.M. White, Degradation pathways for monoethanolamine

in a CO2 capture facility. Energy & Fuels, 2003. 17(4): p. 1034-1039. 45. Dandeneau, R.D. and E. Zerenner, An investigation of glasses for capillary chromatography.

Journal of High Resolution Chromatography, 1979. 2(6): p. 351-356.

58

46. Zechmeister, L., et al., Principles and Practice of Chromatography. Principles and Practice of Chromatography., 1943(2nd Edit).

47. Silverstein, R. and F. Webster, Spectrometric Identification of Organic Compounds6. 2006: John Wiley & Sons.

48. Hoffmann, E., Mass spectrometry. 1996: Wiley Online Library. 49. Revesz, K.M., J.M. Landwehr, and J. Keybl, Measurement of delta13C and delta18O Isotopic

Ratios of CaCO3 Using a Thermoquest Finnigan GasBench II Delta Plus XL Continuous Flow Isotope Ratio Mass Spectrometer With Application to Devils Hole Core DH-11 Calcite. 2001, DTIC Document.

50. Lepaumier, H., et al., Degradation of MMEA at absorber and stripper conditions. Chemical Engineering Science, 2011. 66(15): p. 3491-3498.

51. Ma’mun, S., et al., Selection of new absorbents for carbon dioxide capture. Energy Conversion and Management, 2007. 48(1): p. 251-258.

52. Kjeldahl, J., A new method for the determination of nitrogen in organic matter. Z. Anal. Chem, 1883. 22: p. 366.

53. Yeh, J.T., H.W. Pennline, and K.P. Resnik, Study of CO2 absorption and desorption in a packed column. Energy & fuels, 2001. 15(2): p. 274-278.

54. Poole, C.F. and S.K. Poole, Chromatography today. 1991: Elsevier Science Publishers. 55. Johansen, M.T., Degradation of absorbent systems, in Department of Chemistry. 2012,

Norwegian University of Science and Technology: Trondheim. 56. Eide-Haugmo, I., Environmental impacts and aspects of absorbents used for CO2 capture.

2011, Norwegian University of Science and Technology. 57. Davis, J. and G. Rochelle, Thermal degradation of monoethanolamine at stripper conditions.

Energy Procedia, 2009. 1(1): p. 327-333. 58. Knuutila, H., et al., Formation and destruction of NDELA in 30wt% MEA (monoethanolamine)

and 50wt% DEA (diethanolamine) solutions. Oil and Gas Science and Technology Journal, 2013.

I

Appendix A: External standards used for quantification of degradation

products by GS-MS and IC-EC Table A.1: Standards used for quantification on a GC-MS apparatus.

Standard Abb. Purity [%] CAS No. Lot No. Supplier

N-(2-hydroxyethyl) formamine

HEF 97 693-06-1 G28W033 Alfa Aesar

N, N’-bis(2-hydroxyethyl) oxalamide

BHEOX 99 1871-89-2 E4544B Alfa Aesar

N-(2-hydroxyethyl) imidazolidinone

HEIA 97 3699-54-5 J26U041 Alfa Aesar

2-oxazolidinone OZD 98 497-25-6 MKBJ2536V Aldrich

N-(2-hydroxyethyl) imidazole HEI 97 1615-14-1 1420DH Aldrich

4-(2-hydroxyethyl) piperazin-2-one

HEPO 97 23936-04-1 H30070 Tyger

N-(2-hydroxyethyl) acetamine HEA 99 142-26-7 100455 Aldrich

Table A.2: Standards used for quantification by IC-EC apparatus.

Standard Purity [%] CAS No. Lot No. Supplier

Oxalic acid 99 144-62-7 BCBO6467V Sigma-Aldrich

Formic acid 98 64-18-6 1434094 Fluka analytical

Sodium nitrite 97 7632-00-0 1451880 Sigma-Aldrich

Potassium nitrite 99 7757-79-1 06228CJ Sigma-Aldrich

Glycolic acid 99 79-14-1 2.84E+08 Fluka analytical

Sodium sulfite 99 7757-82-6 A019196201 Sigma-Aldrich

II

Appendix B1: Calibration of mass flow controllers

Calibration of MFC 1

Innstilt gassflow Målt mengde gass Tid

Middel (NL/min)

(%) av max (L)

(s) (L/min) (L/min) omregnet

30 0.027 14.10 0.115

0.027

14.10 0.115

0.027

14.09 0.115

0.027

14.15 0.114 0.115 0.107

20 0.027 21.03 0.077

0.027

21.00 0.077

0.027

21.03 0.077

0.027

21.09 0.077 0.077 0.072

10

0.027 41.75 0.039

0.027

41.72 0.039

0.027

41.78 0.039

0.027 41.72 0.039 0.039 0.036

2

0.004 24.94 0.008

0.004

24.66 0.009

0.004

24.79 0.008

0.004 24.62 0.009 0.008 0.008

Figure B1-1: MFC 1: Values found for calibration

III


Figure B1-2: MFC 2: Values found for calibration


Middel (NL/min)

(%) av max (L)


60 1.000 17.06 3.517

1.000

16.97 3.536

1.000

17.00 3.529

1.000

16.97 3.536 3.529 3.278

30 0.500 16.26 1.845

0.500

16.25 1.846

0.500

16.28 1.843

0.500

16.25 1.846 1.845 1.714

10

0.200 16.50 0.727

0.200

16.47 0.729

0.200

16.59 0.723

0.200 16.50 0.727 0.727 0.675

IV


MFC 3: Values found for calibration


Middel (NL/min)

(%) av max (L)


30 0.027 13.38 0.121

0.027

13.31 0.122

0.027

13.31 0.122

0.027

13.41 0.121 0.121 0.113

10 0.027 35.54 0.046

0.027

35.56 0.046

0.027

35.54 0.046

0.027

35.56 0.046 0.046 0.042

2

0.004 13.69 0.015

0.004

13.69 0.015

0.004

13.66 0.015

0.004 13.71 0.015 0.015 0.014

V




Middel (NL/min)

(%) av max (L)


60 1.000 17.75 3.380

1.000

17.69 3.392

1.000

17.78 3.375

1.000

17.72 3.386 3.383 3.143

30 0.500 17.59 1.706

0.500

17.54 1.710

0.500

17.50 1.714

0.500

17.47 1.717 1.712 1.590

10

0.200 19.88 0.604

0.200

19.88 0.604

0.200

20.04 0.599

0.200 19.91 0.603 0.602 0.559

VI

Calibration of MFC : First calibration

MFC 5: Values found for calibration; Calibration 1


Middel (NL/min)

(%) av max (L)


50 0.50 23.59 1.27

0.50

23.50 1.28

0.50

23.44 1.28

1.28 1.14

30 0.50 39.50 0.76

0.50

39.54 0.76

0.50

39.56 0.76

0.76 0.68

10

0.20 47.22 0.25

0.20

47.35 0.25

0.20

47.13 0.25

0.20 47.31 0.25 0.25 0.23

5

0.10

48.00 0.13

0.10

48.03 0.12

0.10

48.50 0.12

0.10 0.12 0.11

VII

Calibration of MFC : Second calibration

MFC 5: Values found for calibration; Calibration 2


Middel (NL/min)

(%) av max (L)


50 0.50 23.03 1.30

0.50

23.06 1.30

0.50

23.08 1.30

0.50

23.10 1.30 1.30 1.19

30 0.20 15.44 0.78

0.20

15.62 0.77

0.20

15.52 0.77

0.20

15.56 0.77 0.77 0.71

10

0.10 23.25 0.26

0.10

23.44 0.26

0.10

23.62 0.25

0.10 23.44 0.26 0.26 0.23

VIII




Middel (NL/min)

(%) av max (L)


30 0.03 20.47 0.08

0.03

20.50 0.08

0.03

20.34 0.08 0.08 0.07

20 0.03 29.91 0.05

0.03

29.69 0.05

0.03

29.85 0.05 0.05 0.05

10

0.03 62.63 0.03

0.03

62.56 0.03

0.03 62.53 0.03 0.03 0.02

IX




Middel (NL/min)

(%) av max (L)


60 0.800 24.24 1.980

0.800

24.28 1.977

0.800

24.28 1.977

0.800

24.29 1.976 1.978 1.837

40 0.500 22.91 1.309

0.500

22.94 1.308

0.500

22.92 1.309

0.500

22.93 1.308 1.309 1.216

20

0.200 15.60 0.769

0.200

15.50 0.774

0.200

15.40 0.779

0.200 15.50 0.774 0.774 0.719

10

0.100 19.19 0.313

0.100

19.16 0.313

0.100

19.10 0.314

0.100 19.10 0.314 0.314 0.291

X

Appendix B2: Mass flow controller parameters for experiments

Table B1: Openings and flows through MFCs for experiment M1 and M2

MFC's for CO2 stream Composition Flow Opening of flow controller

vol% NL/min %

Reaktor 1 (MFC 1) 100 0.009 3.40

Reaktor 2 (MFC 2) 100 0.0073 3.40

MFC's for N2/O2 stream Composition Flow Opening of flow controller

vol% NL/min %

Reactor 1 (MFC 2) 100 0.387 6.80

Reactor 2 (MFC 4) 100 0.334 6.31

MFC's for mixing gas Composition Flow Opening of flow controller

vol% NL/min %

O2 (MFC 6) 6.00 0.38 19.99

N2 (MFC 5) 94.00 0.38 30.15

Total 100 0.76



vol% NL/min %

Reaktor 1 (MFC 1) 100 0.0075 2.81

Reaktor 2 (MFC 2) 100 0.0075 3.50

MFC's for mixed stream Composition Flow Opening of flow controller

vol% NL/min %

O2 + N2 (MFC 2) 100 0.35 6.13

O2 + N2 (MFC 4) 100 0.35 6.61


vol% NL/min %

O2 (MFC 7) 50.00 0.38 12.63

N2 (MFC 5) 50.00 0.38 16.04

100 0.76

XI



vol% NL/min %

Reaktor 1 (MFC 1) 100 0.0075 2.81

Reaktor 2 (MFC 2) 100 0.0075 3.50

MFC's for mixed stream Composition Flow Opening of flow controller

vol% NL/min %

O2 + N2 (MFC 2) 100 0.35 6.13

O2 + N2 (MFC 4) 100 0.35 6.61


vol% NL/min %

O2 (MFC 7) 6.01 0.0454 1.73

N2 (MFC 5) 93.99 0.7106 29.93

100 0.756

XII

Appendix C : Measured parameters for gas-washing bottles

Tables C1: Experiment M1: 75 °C, 98% O2, Measured H2SO4 and NH3 in samples

Gas-washing bottle 1

Sample nr. P.nr Date Time Day 0.5 M H2SO4 (g) added Mass of H2SO4 before sam-pling [g]

1 P13599 07.02.13 13:45 3.0 94.2 79.31

2 P13600 12.02.13 14:40 8.0 75.8 45.41

3 P13601 17.02.13 14:10 13.0 143.2 126.95

4 P13602 21.02.13 18:50 17.2 120.6 105.81

5 P13603 25.02.13 17:20 21.1 150.5 132.76



1 P13604 07.02.13 13:45 3.0 46.4 46.32

2 P13605 12.02.13 14:40 8.0 43.1 42.86

3 P13606 17.02.13 14:10 13.0 68.2 68.12

4 P13607 21.02.13 18:50 17.2 65.1 65.01

5 P13608 25.02.13 17:20 21.1 85.5 85.00


Sample nr. P.nr Date Time Day NH3 (ug/mL) NH3 ug

1 P13599 07.02.13 13:45 3.0 28736 2279084.677

2 P13600 12.02.13 14:40 8.0 46414 2107678.358

3 P13601 17.02.13 14:10 13.0 11044 1402020.566

4 P13602 21.02.13 18:50 17.2 22883 2421249.172

5 P13603 25.02.13 17:20 21.1 5082 674626.578



1 P13604 07.02.13 13:45 3.0 86 3966.597451

2 P13605 12.02.13 14:40 8.0 33956 1455348.417

3 P13606 17.02.13 14:10 13.0 3 210.3552412

4 P13607 21.02.13 18:50 17.2 1986 129127.7378

5 P13608 25.02.13 17:20 21.1 1 92.6058

XIII




1 P13621 07.02.13 13:55 3.0 98.1 83.22

2 P13622 12.02.13 14:50 8.0 79.9 50.94

3 P13623 17.02.13 14:30 13.0 147.1 118.16

4 P13624 21.02.13 19:00 17.2 115.0 91.53

5 P13625 25.02.13 17:40 21.2 147.4 126.87

6 P13626 06.03.13 12:25 29.9 123.8 75.29

7 P13627 12.03.13 13:00 36.0 150.9 115.44



1 P13628 07.02.13 13:55 3.0 55.5 54.47

2 P13629 12.02.13 14:50 8.0 51.2 49.26

3 P13630 17.02.13 14:30 13.0 79.2 76.92

4 P13631 21.02.13 19:00 17.2 73.7 71.77

5 P13632 25.02.13 17:40 21.2 80.2 78.73

6 P13633 06.03.13 12:25 29.9 75.6 71.74

7 P13634 12.03.13 13:00 36.0 84.6 82.15



1 P13621 07.02.13 13:55 3.0 4481 372914.9533

2 P13622 12.02.13 14:50 8.0 7284 371052.5532

3 P13623 17.02.13 14:30 13.0 5946 702533.7975

4 P13624 21.02.13 19:00 17.2 13195 1207737.435

5 P13625 25.02.13 17:40 21.2 4802 609202.1965

6 P13626 06.03.13 12:25 29.9 24389 1836221.459

7 P13627 12.03.13 13:00 36.0 4635 535012.7175



1 P13628 07.02.13 13:55 3.0 < 1

2 P13629 12.02.13 14:50 8.0 8.0 394.08

3 P13630 17.02.13 14:30 13.0 < 1

4 P13631 21.02.13 19:00 17.2 4.0 287.08

5 P13632 25.02.13 17:40 21.2 < 1

6 P13633 06.03.13 12:25 29.9 14.0 1004.36

7 P13634 12.03.13 13:00 36.0 < 1

XIV




1 P13843 21.03.13 16:00 7.0 138.4 82.74

2 P13844 28.03.13 11:40 13.8 155.6 106.39

3 P13845 04.04.13 16:30 21.0 160.0 106.08

4 P13846 11.04.13 15:40 28.0 157.3 128.86



1 P13847 21.03.13 16:00 7.0 71.2 70.15

2 P13848 28.03.13 11:40 13.8 75.9 74.64

3 P13849 04.04.13 16:30 21.0 69.1 67.58

4 P13850 11.04.13 15:40 28.0 78.7 60.00



1 P13843 21.03.13 16:00 7.0 32701 2705641.025

2 P13844 28.03.13 11:40 13.8 27663 2943117.637

3 P13845 04.04.13 16:30 21.0 22517 2388564.11

4 P13846 11.04.13 15:40 28.0 187 24146.26358



1 P13847 21.03.13 16:00 7.0 2901 203538.1205

2 P13848 28.03.13 11:40 13.8 467 34870.40477

3 P13849 04.04.13 16:30 21.0 8 524.4208

4 P13850 11.04.13 15:40 28.0 < 1

XV




1 P13859 21.03.13 16:00 7.0 146.2 101.37

2 P13860 28.03.13 11:40 13.8 157.9 128.85

3 P13861 04.04.13 16:30 21.0 169.9 124.64

4 P13862 11.04.13 15:40 28.0 153.8 107.91



1 P13863 21.03.13 16:00 7.0 82.0 78.68

2 P13864 28.03.13 11:40 13.8 84.6 82.49

3 P13865 04.04.13 16:30 21.0 75.9 72.33

4 P13866 11.04.13 15:40 28.0 75.7 72.03



1 P13859 21.03.13 16:00 7.0 28907 2930282.316

2 P13860 28.03.13 11:40 13.8 11233 1447411.994

3 P13861 04.04.13 16:30 21.0 24817 3093205.837

4 P13862 11.04.13 15:40 28.0 24201 2611552.571



1 P13863 21.03.13 16:00 7.0 2091 164499.2658

2 P13864 28.03.13 11:40 13.8 7 552.8372563

3 P13865 04.04.13 16:30 21.0 958 69261.61674

4 P13866 11.04.13 15:40 28.0 14 977.2540596

XVI




1 P131075 23.04.13 13:20 7.0 137.7 84.37

2 P131076 30.04.13 12:50 14.0 141.8 89.58

3 P131077 07.05.13 11:55 20.9 153.4 117.26

4 P131078 14.05.13 15:40 28.1 146.5 89.87

5 P131079 21.05.13 10:30 34.9 158.9 105.62

6 P131080 28.05.13 09:40 41.8 136.5 82.98



1 P131081 23.04.13 13:20 7.0 85.0 84.03

2 P131082 30.04.13 12:50 14.0 74.9 73.89

3 P131083 07.05.13 11:55 20.9 83.1 82.47

4 P131084 14.05.13 15:40 28.1 77.0 75.95

5 P131085 21.05.13 10:30 34.9 80.5 79.30

6 P131086 28.05.13 09:40 41.8 79.8 78.48



1 P131075 23.04.13 13:20 7.0 3827 322885.8461

2 P131076 30.04.13 12:50 14.0 5054 452709.6398

3 P131077 07.05.13 11:55 20.9 2451 287363.5708

4 P131078 14.05.13 15:40 28.1 6726 604481.8865

5 P131079 21.05.13 10:30 34.9 4049 427690.3402

6 P131080 28.05.13 09:40 41.8 5229 433877.609



1 P131081 23.04.13 13:20 7.0 0

2 P131082 30.04.13 12:50 14.0 0

3 P131083 07.05.13 11:55 20.9 0

4 P131084 14.05.13 15:40 28.1 0

5 P131085 21.05.13 10:30 34.9 0

6 P131086 28.05.13 09:40 41.8 0

XVII




1 P131096 23.04.13 13:20 7.0 130.9 84.82

2 P131097 30.04.13 12:50 14.0 150.1 105.41

3 P131098 07.05.13 11:55 20.9 154.7 124.42

4 P131099 14.05.13 15:40 28.1 147.3 96.25

5 P131100 21.05.13 10:30 34.9 128.8 80.72

6 P131101 28.05.13 09:40 41.8 144.1 93.83



1 P131102 23.04.13 13:20 7.0 78.7 75.22

2 P131103 30.04.13 12:50 14.0 74.4 70.84

3 P131104 07.05.13 11:55 20.9 84.2 81.86

4 P131105 14.05.13 15:40 28.1 74.6 70.49

5 P131106 21.05.13 10:30 34.9 75.2 71.58

6 P131107 28.05.13 09:40 41.8 80.1 76.42



1 P131096 23.04.13 13:20 7.0 7834 664451.7198

2 P131097 30.04.13 12:50 14.0 8304 875283.2139

3 P131098 07.05.13 11:55 20.9 3685 458451.1205

4 P131099 14.05.13 15:40 28.1 7369 709231.6963

5 P131100 21.05.13 10:30 34.9 7812 630604.4164

6 P131101 28.05.13 09:40 41.8 6639 622977.5292



1 P131102 23.04.13 13:20 7.0 2 P131103 30.04.13 12:50 14.0 3 P131104 07.05.13 11:55 20.9 4 P131105 14.05.13 15:40 28.1 5 P131106 21.05.13 10:30 34.9 6 P131107 28.05.13 09:40 41.8

XVIII

Appendix D: Results for quantification of compounds in solution

D1: Experiment M1: 75 °C, 98% O2

Table D1: LC-MS results for ammonia and primary/secondary amines for samples from reactor.

Time (Days)

NH3 Unit

Time (Days)

Methyl-amine

Ethyl-amine

Dimethyl-amine

Diethyl-amine

Unit

0 97 µg/ml

0 113 < 100 50 449 ng/ml

0.9 1104 µg/ml

0.9 348 < 100 40 < 100 ng/ml

3 1546 µg/ml

3 521 < 100 29 < 100 ng/ml

6.4 1843 µg/ml

6.4 660 < 100 19 < 100 ng/ml

8.1 1679 µg/ml

8.1 655 < 100 18 < 100 ng/ml

10.1 1949 µg/ml

10.1 718 < 100 18 < 100 ng/ml

13 1891 µg/ml

13 719 < 100 14 < 100 ng/ml

17.2 1626 µg/ml

17.2 736 < 100 < 10 < 100 ng/ml

21.2 1413 µg/ml

21.2 753 < 100 11 < 100 ng/ml

Table D2: LC-MS results for degradation mix compounds in samples from reactor.

Time (Days)

HeGly HEF BHEOX HEA HEPO OZD HEI Unit

0 10 159 11 4 < 1 < 10 20 µg/mL

0.9 581 6616 637 75 20 147 3086 µg/mL

3 1324 13117 982 319 155 620 7506 µg/mL

6.4 1404 14108 797 610 330 998 9264 µg/mL

8.1 1534 13838 671 720 369 1076 9286 µg/mL

10.1 1561 13071 655 847 394 1155 9413 µg/mL

13 1637 12772 548 991 452 1154 9625 µg/mL

17.2 1974 12704 518 1204 551 1304 10393 µg/mL

21.2 1892 11319 424 1190 585 1187 9463 µg/mL

Table D3: Concentration of MEA and DEA in samples from reactor found by LC-MS analysis.

Time (Days) MEA Unit Time (Days) DEA Unit

0 4.62 mol/L

0 0.07 mmol/L

0.9 4.7 mol/L

0.9 0.29 mmol/L

3 3.05 mol/L

3 0.41 mmol/L

6.4 2.29 mol/L

6.4 0.38 mmol/L

8.1 1.91 mol/L

8.1 0.4 mmol/L

10.1 1.72 mol/L

10.1 0.42 mmol/L

13 1.62 mol/L

13 0.45 mmol/L

17.2 1.72 mol/L

17.2 0.5 mmol/L

21.2 1.52 mol/L

21.2 0.55 mmol/L

XIX

Table D5: Concentration of anions in samples from reactor analyzed by IC-EC

Time (Days) Formate (PPM) Oxalate (PPM) Nitrate (PPM) Sulphate (PPM) Nitrite (PPM)

0 39 #N/A #N/A 1 067 #N/A

1 909 101 92 1 096 717

3 7 159 520 334 1 295 1 885

6 #N/A 1 280 953 1 264 3 608

8 20 168 1 797 1 013 1 211 3 779

10 #N/A 2 225 1 072 1 112 3 746

13 #N/A 2 929 1 055 1 135 3 892

17 25 016 4 115 1 879 1 188 3 428

21 25 579 5 388 1 878 1 152 3 379

Table D6: Amine and CO2 concentration found for samples from the reactor by titration.

Time (days)

Amin (mol/kg) CO2 (mol/kg)

0.0 0.00 1.78

0.9 0.95

3.0 3.03

6.4 6.38

8.1 8.07

10.1 10.11

13.0 13.05

17.2 17.24

21.2 21.18 0.11

Table D7: Parameters and concentrations found of HEF and HEI in run in parallels in GC-MS

Sample F1P1 F1P2 F1P3 F1P4 F1P5 F1P6 F1P7 F1P8 F1P9

Time (Days) 0.0 1.0 3.0 6.4 8.1 10.1 13.0 17.2 21.2

Methanol (g) 0.84 0.80 0.80 0.80 0.79 0.79 0.80 0.79 0.79

Sample (g) 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02

Density (g/mL) 1.09 1.09 1.09 1.09 1.08 1.08 1.08 1.07 1.07

Total (g) 0.86 0.82 0.82 0.82 0.81 0.81 0.82 0.81 0.81

HEF 1 diluted 58.0 396.5 852.5 571.4 680.6 615.0 524.2 564.5

HEF 2 diluted

126.5 446.5 573.1 617.3 757.2 753.0 622.3 728.0

HEF 3 diluted

800.6

HEF 4 diluted 383.6

HEPO 1 diluted 10.0 19.9 36.5 47.2 37.7 39.2 34.8 27.4

HEPO 2 diluted

20.1 25.2 22.8

58.2 57.0 39.7 45.4

HEPO 3 diluted

43.4

HEPO 4 diluted 28.5

Avr. HEF 92.2 421.5 712.8 594.3 718.9 684.0 573.2 619.2

STD (%) 52.6 8.4 27.7 5.5 7.5 14.3 12.1 30.0

Avr. HEPO 15.1 22.6 29.6 47.2 48.0 48.1 37.2 36.2

STD (%) 47.7 16.6 32.8 30.2 26.3 9.3 26.4

XX



Time (days) NH3 Unit

Methyl-amine

Ethyl-amine

Dimethyl-amine

Diethyl-amine

Unit

0 97 µg/ml

< 10 358 17 2933 ng/ml

0.9 167 µg/ml

42 244 20 1326 ng/ml

3 213 µg/ml

150 265 24 770 ng/ml

6.3 213 µg/ml

392 284 42 569 ng/ml

7.8 212 µg/ml

466 250 36 402 ng/ml

10.1 207 µg/ml

559 233 42 351 ng/ml

13 218 µg/ml

749 262 50 292 ng/ml

17.2 244 µg/ml

1089 302 61 428 ng/ml

21.2 253 µg/ml

1325 256 65 245 ng/ml

24 283 µg/ml

1527 241 70 208 ng/ml

29.9 331 µg/ml

1845 207 67 170 ng/ml

35.9 383 µg/ml

2126 232 68 186 ng/ml


Time (days) HeGly HEF BHEOX HEA HEPO OZD HEI Unit

0 5 157 10 3 < 1 < 10 18 µg/mL

0.9 202 418 21 10 3 < 10 140 µg/mL

3 1217 816 43 32 13 24 591 µg/mL

6.3 3294 1348 66 70 39 39 1198 µg/mL

7.8 4307 1523 64 84 53 43 1399 µg/mL

10.1 4989 1697 65 99 69 45 1526 µg/mL

13 6322 1945 85 148 102 59 1818 µg/mL

17.2 7174 2474 98 197 145 78 2071 µg/mL

21.2 7422 2692 105 242 191 82 2172 µg/mL

24 7733 3062 113 295 222 99 2317 µg/mL

29.9 7745 3423 125 365 312 128 2469 µg/mL

35.9 7356 3931 145 465 407 165 2561 µg/mL

XXI


Time (days) MEA Unit

DEA Unit

0.0 5.27 mol/L

0.07 mmol/L

0.9 5.28 mol/L

0.11 mmol/L

3.0 4.95 mol/L

0.15 mmol/L

6.3 4.72 mol/L

0.2 mmol/L

7.8 4.91 mol/L

0.22 mmol/L

10.1 4.79 mol/L

0.24 mmol/L

13.0 4.43 mol/L

0.29 mmol/L

17.2 4.12 mol/L

0.36 mmol/L

21.2 3.93 mol/L

0.37 mmol/L

24.0 3.98 mol/L

0.44 mmol/L

29.9 3.78 mol/L

0.5 mmol/L

35.9 3.47 mol/L

0.54 mmol/L



0 #N/A 64 #N/A 1 321 #N/A

1 71 58 #N/A 1 356 #N/A

3 303 60 #N/A 1 266 #N/A

6 702 88 #N/A 1 340 #N/A

8 1 001 119 #N/A 1 360 44

10 1 132 148 #N/A 1 275 45

13 1 649 195 17 1 267 63

17 2 374 346 #N/A 1 258 74

21 3 215 461 34 1 295 90

24 3 939 579 36 1 264 103

30 5 345 798 67 1 257 145

36 6 109 1 650 81 2 414 173

XXII


Time (days) Amin (mol/kg) CO2 (mol/kg)

0.0 4.51 1.77

0.9 4.45

3.0 4.33

6.3 4.12

7.8 4.09

10.1 3.93

13.0 3.79

17.2 3.69

21.2 3.46

24.0 3.37

29.9 3.18

35.9 2.98 1.06

XXIII



Time (Days)

NH3 Unit

Methyl-amine

Ethyl-amine

Dimethyl-amine

Diethyl-amine

Unit

0 45 µg/ml

< 10 1063 19 8717 ng/ml

1 383 µg/ml

55 582 18 3809 ng/ml

3 498 µg/ml

140 431 13 3048 ng/ml

7 580 µg/ml

221 166 < 10 1451 ng/ml

10 616 µg/ml

280 135 < 10 1042 ng/ml

14 745 µg/ml

308 < 100 < 10 504 ng/ml

21 946 µg/ml

359 < 100 < 10 284 ng/ml

28 1854 µg/ml

409 < 100 < 10 218 ng/ml



0 5 65 3 5 < 1 < 10 5 µg/mL

1 170 2381 263 19 4 75 367 µg/mL

3 630 4694 555 72 31 271 1007 µg/mL

7 1214 7357 740 197 132 639 1819 µg/mL

10 1358 8649 827 315 211 947 2325 µg/mL

14 1369 9745 793 451 277 1227 2775 µg/mL

21 1306 10431 750 673 346 1611 3290 µg/mL

28 1313 10227 679 821 294 1926 3816 µg/mL


Time (Days)

MEA Unit

DEA Unit

0 5.26 mol/L

0.06 mmol/L

1 4.51 mol/L

0.13 mmol/L

3 4.48 mol/L

0.20 mmol/L

7 4.00 mol/L

0.27 mmol/L

10 3.58 mol/L

0.28 mmol/L

14 3.37 mol/L

0.30 mmol/L

21 2.69 mol/L

0.32 mmol/L

28 2.36 mol/L

0.37 mmol/L

XXIV


Time (days) Amin (mol/kg) CO2 (mol/kg)

0.0 4.47 1.78

1.1 4.29

3.2 3.97

7.1 3.39

10.0 3.09

14.0 2.65

21.2 2.01

28.1 1.73 0.68



0 #N/A #N/A #N/A #N/A #N/A

1 273 #N/A 42 #N/A 249

3 1 083 64 105 #N/A 596

7 3 042 266 243 #N/A 1 073

10 #N/A #N/A #N/A #N/A #N/A

14 8 096 908 507 #N/A 1 654

21 10 349 1 768 678 #N/A 2 074

28 9 975 2 663 968 #N/A 2 191

XXV



Time (days)

NH3 Unit

Methyl-amine

Ethyl-amine

Dimethyl-amine

Diethyl-amine

Unit

0 517 µg/ml

171 < 100 29 817 ng/ml

1.1 40 µg/ml

23 < 100 27 554 ng/ml

3.2 737 µg/ml

382 < 100 25 457 ng/ml

7.1 914 µg/ml

627 < 100 19 218 ng/ml

10 1243 µg/ml

687 < 100 16 135 ng/ml

14 936 µg/ml

1085 < 100 43 1527 ng/ml

21.2 1540 µg/ml

740 < 100 15 142 ng/ml

28.1 1438 µg/ml

733 < 100 14 115 ng/ml


Time (days)


0 2 50 2 < 1 < 1 < 10 3 µg/mL

1 495 2154 236 39 14 93 483 µg/mL

3 1585 4539 434 171 121 353 1514 µg/mL

7 2005 6950 571 469 331 843 2566 µg/mL

10 1550 8517 677 653 327 1236 3150 µg/mL

14 1897 8390 582 947 528 1338 3830 µg/mL

21 1296 9478 542 1273 312 2066 4055 µg/mL


Time (days)

MEA Unit

DEA OZD

0 5.43 mol/L

0.25 < 10

1 5.06 mol/L

0.06 < 10

3 4.17 mol/L

0.4 24

7 3.44 mol/L

0.38 39

10 3.14 mol/L

0.29 43

14 3.16 mol/L

0.36 45

21 2.06 mol/L

0.21 59

28 1.58 mol/L

0.21 78

XXVI


Time (days)

Amin (mol/kg)

CO2 (mol/kg)

0.0 4.48 1.76

1.1 4.28

3.2 3.86

7.1 3.09

10.0 2.55

14.0 2.18

21.2 1.21

28.1 0.72 0.21



0 #N/A #N/A #N/A #N/A #N/A

1 389 #N/A 31 #N/A 151

3 1 632 121 92 #N/A 435

7 4 875 582 244 #N/A 850

10 8 770 1 195 385 #N/A 1 224

14 #N/A 1 953 460 #N/A 1 128

21 #N/A 4 132 781 #N/A 1 536

28 22272.29372 6 137 1 034 #N/A 1 271

XXVII



Time (days)

MEA Unit

DEA Unit

0 5.56 mol/L

0.034 mmol/L

1 5.71 mol/L

0.056 mmol/L

3 5.46 mol/L

0.081 mmol/L

7 5.33 mol/L

0.132 mmol/L

14 5.03 mol/L

0.191 mmol/L

21 4.93 mol/L

0.213 mmol/L

28 4.59 mol/L

0.243 mmol/L

35 4.66 mol/L

0.281 mmol/L

42 4.49 mol/L

0.302 mmol/L



0 8 104 < 100 < 10 < 1 < 10 3 µg/mL

1 14 228 < 100 < 10 < 1 < 10 14 µg/mL

3 72 478 < 100 10 1 < 10 32 µg/mL

7 279 977 126 17 6 44 70 µg/mL

14 772 1632 184 43 26 124 134 µg/mL

21 1135 2347 271 78 43 216 178 µg/mL

28 1484 2475 212 103 91 285 231 µg/mL

35 1797 2947 251 148 150 347 280 µg/mL

42 1806 2929 230 174 189 406 309 µg/mL


Time (days) NH3 Unit

Methyl-amine

Ethyl-amine

Dimethyl-amine

Diethyl-amine

Unit

0 30 µg/ml

< 10 353 23 1991 ng/ml

1 82 µg/ml

< 10 213 24 1270 ng/ml

3 101 µg/ml

< 10 209 19 768 ng/ml

7 124 µg/ml

27 218 24 552 ng/ml

14 122 µg/ml

46 164 15 339 ng/ml

21 228 µg/ml

65 129 18 343 ng/ml

28 121 µg/ml

84 < 100 15 368 ng/ml

35 136 µg/ml

104 < 100 12 264 ng/ml

42 155 µg/ml

125 < 100 12 195 ng/ml

XXVIII


Time (days) Amin (mol/kg)

CO2 (mol/kg)

0.0 0.00 1.78

1.0 1.05

3.0 2.96

7.0 6.95

13.9 13.92

20.9 20.87

28.0 28.05

34.8 34.83

41.8 41.77 0.11



0 #N/A #N/A #N/A 889 #N/A

1 #N/A #N/A #N/A 794 #N/A

3 28 #N/A #N/A 795 #N/A

7 129 #N/A #N/A 793 51

14 394 66 #N/A 773 115

21 607 96 22 831 156

28 1 041 172 #N/A 752 197

35 1 433 251 #N/A 742 231

42 1 485 448 100 890 164

XXIX



Time (Days)

NH3 Unit

Methyl-amine

Ethyl-amine

Dimethyl-amine

Diethyl-amine

Unit

0 32 µg/ml

< 10 413 24 1845 ng/ml

1 156 µg/ml

28 284 27 997 ng/ml

3 235 µg/ml

68 217 24 673 ng/ml

7 247 µg/ml

153 204 25 602 ng/ml

14 187 µg/ml

259 159 28 354 ng/ml

21 255 µg/ml

380 186 31 320 ng/ml

28 216 µg/ml

480 220 32 266 ng/ml

35 220 µg/ml

553 175 39 145 ng/ml

42 236 µg/ml

613 185 31 163 ng/ml


Time (Days)


0 5 88 < 100 < 10 < 1 < 10 3 µg/mL

1 61 353 < 100 11 < 1 < 10 48 µg/mL

3 300 677 < 100 20 4 33 166 µg/mL

7 1105 1305 < 100 52 24 82 376 µg/mL

14 2527 1789 < 100 121 87 139 596 µg/mL

21 3496 1835 < 100 153 146 151 716 µg/mL

28 3913 2299 < 100 216 207 188 863 µg/mL

35 4016 2483 108 272 270 210 872 µg/mL

42 4105 2890 169 335 343 242 865 µg/mL


Time (Days)

MEA Unit

DEA Unit

0 5.37 mol/L

0.046 mmol/L

1 4.91 mol/L

0.068 mmol/L

3 4.86 mol/L

0.113 mmol/L

7 4.71 mol/L

0.186 mmol/L

14 4.96 mol/L

0.245 mmol/L

21 4.67 mol/L

0.306 mmol/L

28 4.42 mol/L

0.3 mmol/L

35 4.46 mol/L

0.359 mmol/L

42 4.05 mol/L

0.337 mmol/L

XXX


Time (Days) Amin (mol/kg)

CO2 (mol/kg)

0 4.53 1.76

1 4.44

3 4.39

7 4.30

14 4.13

21 4.03

28 3.85

35 3.76

42 3.65 1.49



0 #N/A #N/A #N/A 801 #N/A

1 #N/A #N/A #N/A 784 #N/A

3 82 #N/A #N/A 764 35

7 291 46 #N/A 778 65

14 582 100 #N/A 753 98

21 1 091 189 #N/A 776 108

28 1 319 290 #N/A 707 118

35 2 563 431 74 867 173

42 2 802 697 65 1 017 165

XXXI

Appendix E: LC-MS degradation mix results for mixing experiments.

Table E1: Acidic conditions

Time (Days)


0.10 9938 < 100 < 1 < 1 < 1 < 10 < 1 µg/mL

0.37 12379 < 100 < 1 < 1 < 1 < 10 < 1 µg/mL

1.26 14873 < 100 < 1 < 1 < 1 < 10 < 1 µg/mL

3.00 16909 < 100 < 1 < 1 < 1 < 10 < 1 µg/mL

Table E2: Basic conditions

Time (Days)


0 8 223 < 100 < 10 < 1 < 10 33 µg/mL

0.01 5 928 < 100 < 10 < 1 < 10 77 µg/mL

0.26 34 3757 < 100 < 10 < 1 < 10 146 µg/mL

1.88 94 11175 < 100 < 10 < 1 < 10 202 µg/mL

Degradation of Amines - COnnecting REpositories · 2016-04-20 · Degradation of Amines. Maren...

Documents

Transcript of Degradation of Amines - COnnecting REpositories · 2016-04-20 · Degradation of Amines. Maren...